Quellenverzeichnis
651 Quellen, alphabetisch, mit anklickbaren Links.
Im Buch steht dasselbe Verzeichnis nach Kapiteln geordnet. Arbeiten mit ★ gehören zum Kanon aus Kapitel 24. Alle Onlinequellen wurden am 18. Oktober 2026 abgerufen, sofern nicht anders angegeben.
Abramson, Josh; Adler, Jonas; Dunger, Jack u. a. (2024): Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630, S. 493–500 (Google DeepMind und Isomorphic Labs; letzte Autoren Demis Hassabis und John M. Jumper), 8. Mai 2024. https://www.nature.com/articles/s41586-024-07487-w Abgerufen am 25. Oktober 2026.
AG Energiebilanzen (2026): Stromerzeugung nach Energieträgern (Strommix) von 1990 bis 2025 (in TWh) Deutschland insgesamt. AG Energiebilanzen, Tabelle (Abgabe September 2026; auf der Datenseite eingestellt am 7.10.2026), 16. September 2026. https://ag-energiebilanzen.de/wp-content/uploads/Strerz-Abgabe-September-2026.pdf https://ag-energiebilanzen.de/daten-und-fakten/zusatzinformationen/ Abgerufen am 2. November 2026.
Agarwal, Rishabh; Singh, Avi u. a. (2024): Many-Shot In-Context Learning. NeurIPS 2024 (Spotlight); Preprint arXiv:2404.11018 (Google DeepMind, über das eigene Modell Gemini 1.5 Pro), 17. April 2024. https://arxiv.org/abs/2404.11018 Abgerufen am 20. Oktober 2026.
Aghajanyan, Armen; Zettlemoyer, Luke; Gupta, Sonal (2021): Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tuning. ACL-IJCNLP 2021; Preprint arXiv:2012.13255 (Facebook; v1 vom 22.12.2020), 22. Dezember 2020. https://arxiv.org/abs/2012.13255 Abgerufen am 24. Oktober 2026.
Ahmed, Ahmed; Cooper, A. Feder; Koyejo, Sanmi u. a. (2026): Extracting books from production language models. Preprint, arXiv:2601.02671, 6. Januar 2026. https://arxiv.org/abs/2601.02671 Abgerufen am 19. Oktober 2026.
AI Security Institute (2026a): Our evaluation of Claude Mythos Preview’s cyber capabilities. AISI, Blog, London, 13. April 2026. https://www.aisi.gov.uk/blog/our-evaluation-of-claude-mythos-previews-cyber-capabilities Abgerufen am 25. Oktober 2026.
AI Security Institute (2026b): Our evaluation of OpenAI’s GPT-5.5 cyber capabilities. AISI, Blog, London, 30. April 2026. https://www.aisi.gov.uk/blog/our-evaluation-of-openais-gpt-5-5-cyber-capabilities Abgerufen am 25. Oktober 2026.
AI Security Institute (2026c): How Far Behind the Frontier are Leading Open Weight Models on Cyber? AISI, Blog, London, 17. Juli 2026. https://www.aisi.gov.uk/blog/how-far-behind-the-frontier-are-leading-open-weight-models-on-cyber Abgerufen am 25. Oktober 2026.
Aleph Alpha (2026a): Kolibri 1, Model Card. Hugging Face, 3. Oktober 2026. https://huggingface.co/Aleph-Alpha/Kolibri-1 https://huggingface.co/Aleph-Alpha/Kolibri-1-BF16 (Herstellerangabe) Abgerufen am 2. November 2026.
Aleph Alpha (2026b): Kolibri: A Sovereign European Model on the Pareto Frontier. Technischer Bericht (PDF, 189 S.), von der Modellkarte verlinkt, 3. Oktober 2026. https://aleph-alpha.com/downloads/tech-report.pdf (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Aleph Alpha (2026c): Kolibri Has Landed: A Sovereign Open-Weight Model. Aleph Alpha, Tech-Blog (Rubrik Research), 3. Oktober 2026. https://aleph-alpha.com/en/blog/kolibri-has-landed-a-sovereign-open-weight-model/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
AlgorithmWatch (o. J.): AlgorithmWatch. Website, Berlin und Zürich. https://algorithmwatch.org/de/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Ali, Mehdi u. a. (2025): Teuken-7B-Base & Teuken-7B-Instruct: Towards European LLMs. Preprint, arXiv:2410.03730 (v1 30.9.2024; v2 15.10.2024 zum Modell v0.4 mit 4 Bio. Tokens; v3 21.8.2025 zu v0.6 mit 6 Bio. Tokens), 21. August 2025. https://arxiv.org/abs/2410.03730 Abgerufen am 19. Oktober 2026.
Altman, Sam (2025a): „insane thing: we are currently losing money on openai pro subscriptions!“ (Beitrag auf X). X, @sama, 6. Januar 2025. https://x.com/sama/status/1876104315296968813 Abgerufen am 26. Oktober 2026.
Altman, Sam (2025b): The Gentle Singularity. Blog blog.samaltman.com, 10. Juni 2025. https://blog.samaltman.com/the-gentle-singularity (Herstellerangabe) Abgerufen am 19. Oktober 2026.
American Dialect Society (2024): American Dialect Society Selects ‚enshittification‘ as 2023 Word of the Year. Pressemitteilung, 5. Januar 2024. https://americandialect.org/wp-content/uploads/2024/01/2023-Word-of-the-Year-PRESS-RELEASE.pdf https://americandialect.org/2023-word-of-the-year-is-enshittification/ Abgerufen am 19. Oktober 2026.
American Film Institute (2013): Her (2013). AFI Catalog of Feature Films, 18. Dezember 2013. https://catalog.afi.com/Catalog/moviedetails/69900 Abgerufen am 2. November 2026.
American Statistical Association (2016): American Statistical Association Releases Statement on Statistical Significance and P-Values (Pressemitteilung); dazu Wasserstein, Ronald L. und Lazar, Nicole A.: The ASA Statement on p-Values: Context, Process, and Purpose. Alexandria (Virginia); The American Statistician 70 (2), S. 129–133, DOI 10.1080/00031305.2016.1154108, 7. März 2016. https://www.amstat.org/asa/files/pdfs/p-valuestatement.pdf https://doi.org/10.1080/00031305.2016.1154108 Abgerufen am 25. Oktober 2026.
Amershi, Saleema; Weld, Daniel S.; Vorvoreanu, Mihaela u. a. (2019): Guidelines for Human-AI Interaction. CHI 2019, Glasgow, DOI 10.1145/3290605.3300233 (13 Autorinnen und Autoren, Microsoft), 29. April 2019. https://doi.org/10.1145/3290605.3300233 Abgerufen am 25. Oktober 2026.
Aminabadi, Reza Yazdani; Rajbhandari, Samyam; Zhang, Minjia u. a. (2022): DeepSpeed Inference: Enabling Efficient Inference of Transformer Models at Unprecedented Scale. SC22; Preprint arXiv:2207.00032 (11 Autoren), 30. Juni 2022. https://arxiv.org/abs/2207.00032 Abgerufen am 19. Oktober 2026.
Amodei, Dario (2025): The Urgency of Interpretability. Essay, Website von Dario Amodei. https://www.darioamodei.com/post/the-urgency-of-interpretability (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Amtsblatt der Europäischen Union (2025): Appeal brought on 31 October 2025 by Philippe Latombe against the judgment of the General Court (Tenth Chamber, Extended Composition) delivered on 3 September 2025 in Case T-553/23, Latombe v Commission (Case C-703/25 P). ABl. C/2025/6610, 22. Dezember 2025. Stand laut InfoCuria (Fallregister des Gerichtshofs) am 2.11.2026 anhängig. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ:C_202506610 Abgerufen am 19. Oktober 2026.
Amtsgericht Köln (2025): Beschluss 312 F 130/25. ECLI:DE:AGK:2025:0702.312F130.25.00; Justizportal NRW (NRWE), 2. Juli 2025. https://nrwe.justiz.nrw.de/ag_koeln/j2025/312_F_130_25_Beschluss_20250702.html Abgerufen am 19. Oktober 2026.
An, Jiafu; Huang, Difang; Lin, Chen u. a. (2025): Measuring gender and racial biases in large language models: Intersectional evidence from automated resume evaluation. PNAS Nexus 4 (3), pgaf089 (online 12.3.2025, Ausgabe vom 27.2.2025), 12. März 2025. https://doi.org/10.1093/pnasnexus/pgaf089 Abgerufen am 25. Oktober 2026.
Andreas, Jacob; Van den Broeck, Guy; Risteski, Andrej u. a. (2026): Submission policies for ICLR 2027. ICLR Blog, 2. September 2026. https://blog.iclr.cc/2026/09/02/submission-policies-for-iclr-2027/ Abgerufen am 19. Oktober 2026.
Anthropic (2023a): Claude’s constitution. Anthropic, News, 9. Mai 2023. https://www.anthropic.com/news/claudes-constitution (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2023b): Collective Constitutional AI: Aligning a Language Model with Public Input. Anthropic, Research (gemeinsam mit dem Collective Intelligence Project), 17. Oktober 2023. https://www.anthropic.com/research/collective-constitutional-ai-aligning-a-language-model-with-public-input (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2024a): Many-shot jailbreaking. Anthropic, Research (mit dem Paper von Anil, Cem u. a.), 2. April 2024. https://www.anthropic.com/research/many-shot-jailbreaking https://www-cdn.anthropic.com/af5633c94ed2beb282f6a53c595eb437e8e7b630/Many_Shot_Jailbreaking__2024_04_02_0936.pdf (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2024b): Mapping the Mind of a Large Language Model. Anthropic, Research, 21. Mai 2024. https://www.anthropic.com/research/mapping-mind-language-model (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2024c): Golden Gate Claude. Anthropic, Research, 23. Mai 2024. https://www.anthropic.com/research/golden-gate-claude (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2024d): Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku. Anthropic News, 22. Oktober 2024. https://www.anthropic.com/news/3-5-models-and-computer-use (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2024e): A statistical approach to model evaluations. Anthropic Research (Blog), 19. November 2024. https://www.anthropic.com/research/statistical-approach-to-model-evals (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2024f): Introducing the Model Context Protocol. Anthropic News, 25. November 2024. https://www.anthropic.com/news/model-context-protocol (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2024g): Building effective agents (Erik S., Barry Zhang). Anthropic Engineering, 19. Dezember 2024. https://www.anthropic.com/engineering/building-effective-agents (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2025a): Constitutional Classifiers: Defending against universal jailbreaks. Anthropic, Research, 3. Februar 2025, mit Nachtrag zur öffentlichen Demo vom 13. Februar 2025. https://www.anthropic.com/research/constitutional-classifiers (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2025b): System Card: Claude Sonnet 4.5. Anthropic, PDF (September 2025). https://assets.anthropic.com/m/12f214efcc2f457a/original/Claude-Sonnet-4-5-System-Card.pdf (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2025c): Claude’s extended thinking. Anthropic, News, 24. Februar 2025. https://www.anthropic.com/news/visible-extended-thinking (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2025d): Reasoning models don’t always say what they think. Anthropic, Research (Blog), 3. April 2025. https://www.anthropic.com/research/reasoning-models-dont-say-think (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2025e): How people use Claude for support, advice, and companionship. Anthropic, Research, 27. Juni 2025. https://www.anthropic.com/research/how-people-use-claude-for-support-advice-and-companionship (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2025f): Updates to Consumer Terms and Privacy Policy. Anthropic News, 28. August 2025. https://www.anthropic.com/news/updates-to-our-consumer-terms (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2025g): Effective context engineering for AI agents. Anthropic Engineering; verfasst vom Applied AI Team (Prithvi Rajasekaran, Ethan Dixon, Carly Ryan, Jeremy Hadfield), 29. September 2025. https://www.anthropic.com/research/effective-context-engineering-for-ai-agents (Herstellerangabe) Abgerufen am 23. Oktober 2026.
Anthropic (2025h): Expanding our use of Google Cloud TPUs and Services. Anthropic, News, 23. Oktober 2025. https://www.anthropic.com/news/expanding-our-use-of-google-cloud-tpus-and-services (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2025i): Disrupting the first reported AI-orchestrated cyber espionage campaign. Anthropic, News, 13. November 2025. https://www.anthropic.com/news/disrupting-AI-espionage (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026a): Claude Haiku 4.5 (Modellseite). Claude-Dokumentation. https://platform.claude.com/docs/en/models/haiku-4-5/overview (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026aa): Redacted Risk Report August 2026. Anthropic, PDF (Stichtag 15.7.2026), 14. August 2026. https://www.anthropic.com/aug-2026-risk-report (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2026ab): Claude Fable 5.1. Claude-Dokumentation, Modellseite, 1. September 2026. https://platform.claude.com/docs/en/models/fable-5-1/overview (Herstellerangabe) Abgerufen am 29. Oktober 2026.
Anthropic (2026ac): Detecting and countering misuse of AI: September 2026. Threat Intelligence Report, 10. September 2026. https://www.anthropic.com/threat-intelligence-report-september-2026 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2026ad): Claude Opus 5.5 System Card. Anthropic, PDF (22.9.2026), 22. September 2026. https://www-cdn.anthropic.com/fc1b44717c85dc068bc6ba5024219938094694bd/Claude%20Opus%205.5%20System%20Card.pdf (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2026b): Claude Opus 5.5 migration guide. Claude-Dokumentation. https://platform.claude.com/docs/en/models/opus-5-5/migration-guide (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2026c): Claude Opus 5.5 (Modellseite). Claude-Dokumentation. https://platform.claude.com/docs/en/models/opus-5-5/overview (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026d): Computer use tool. Claude-Dokumentation (Release Notes 19.8.2026: allgemein verfügbar). https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool https://platform.claude.com/docs/en/release-notes/overview (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026e): Context windows. Claude-Dokumentation. https://platform.claude.com/docs/en/build-with-claude/context-windows (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026f): Data residency. Claude Platform Docs (ohne Datum). https://platform.claude.com/docs/en/manage-claude/data-residency (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026g): How do usage and length limits work? Claude Help Center. https://support.claude.com/en/articles/11647753-understanding-usage-and-length-limits (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2026h): How long do you store my data?; Is my data used for model training?; How do I change my model improvement privacy settings? Anthropic Privacy Center. https://privacy.claude.com/en/articles/10023548-how-long-do-you-store-my-data https://privacy.claude.com/en/articles/10023580-is-my-data-used-for-model-training https://privacy.claude.com/en/articles/12109829-how-do-i-change-my-model-improvement-privacy-settings (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026i): Measurements for understanding the pace of AI development inside frontier labs. The Anthropic Institute, Website (Daten Stand August 2026). https://www.anthropic.com/institute/measuring-pace-of-ai-development (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2026j): Model deprecations. Claude-Dokumentation. https://platform.claude.com/docs/en/about-claude/model-deprecations (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026k): Models overview. Claude-Dokumentation. Stand 02.11.2026. https://platform.claude.com/docs/en/models/overview (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026l): Plans & Pricing. claude.com (Free, Pro, Max). https://claude.com/pricing (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026m): Pricing. Claude-Dokumentation. https://platform.claude.com/docs/en/about-claude/pricing (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026n): System prompts (Übersicht) und Claude Opus 5.5 system prompts (Eintrag vom 27.10.2026). Claude-Dokumentation, Release Notes. https://platform.claude.com/docs/en/release-notes/system-prompts/overview https://platform.claude.com/docs/en/release-notes/system-prompts/claude-opus-5-5 (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026o): Thinking. Claude-Dokumentation (ohne Datum). https://platform.claude.com/docs/en/build-with-claude/thinking (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026p): Tool use with Claude. Claude-Dokumentation. https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026q): Use incognito chats. Claude Help Center. https://support.claude.com/en/articles/12260368-use-incognito-chats https://support.claude.com/en/articles/11817273 (Herstellerangabe) Abgerufen am 2. November 2026.
Anthropic (2026r): Next-generation Constitutional Classifiers: More efficient protection against universal jailbreaks. Anthropic, Research, 9. Januar 2026. https://www.anthropic.com/research/next-generation-constitutional-classifiers (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026s): Claude’s new constitution. Anthropic, News, 22. Januar 2026. https://www.anthropic.com/news/claude-new-constitution (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026t): Detecting and preventing distillation attacks. Anthropic, News, 23. Februar 2026. https://www.anthropic.com/news/detecting-and-preventing-distillation-attacks (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2026u): Alignment Risk Update: Claude Mythos Preview. Anthropic (redigierte Fassung vom 10.4.2026), 7. April 2026. https://www.anthropic.com/claude-mythos-preview-risk-report (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Anthropic (2026v): Project Glasswing: Securing critical software for the AI era. Anthropic, Website, 7. April 2026. https://www.anthropic.com/glasswing (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2026w): System Card: Claude Mythos Preview. Anthropic, PDF (Changelog vom 8.4.2026), 7. April 2026. https://www-cdn.anthropic.com/08ab9158070959f88f296514c21b7facce6f52bc.pdf (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (2026x): Claude Opus 4.7. Claude-Dokumentation, Modellseite (veröffentlicht 16.4.2026), 16. April 2026. https://platform.claude.com/docs/en/models/opus-4-7/overview (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Anthropic (2026y): Claude Fable 5 and Claude Mythos 5. Anthropic, News, 9. Juni 2026. https://www.anthropic.com/news/claude-fable-5-mythos-5 (Herstellerangabe) Abgerufen am 29. Oktober 2026.
Anthropic (2026z): How Claude’s text watermarking works. Anthropic, News (Aktualisierung vom 1.9.2026); dazu Help Center „How Claude marks AI-generated content“ (11.8.2026), 14. August 2026. https://www.anthropic.com/news/claude-text-watermark (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Anthropic (o. J.): Consumer Terms of Service; Commercial Terms of Service. Anthropic, Legal. https://www.anthropic.com/legal/consumer-terms https://www.anthropic.com/legal/commercial-terms (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Apache Software Foundation (2004): Apache License, Version 2.0. apache.org, Januar 2004. https://www.apache.org/licenses/LICENSE-2.0 Abgerufen am 19. Oktober 2026.
Apollo Research (2025): Stress Testing Deliberative Alignment for Anti-Scheming Training. Apollo Research, Website (in Zusammenarbeit mit OpenAI), 17. September 2025. https://www.apolloresearch.ai/science/stress-testing-deliberative-alignment-for-anti-scheming-training Abgerufen am 25. Oktober 2026.
Apple (2026): Apple introduces new Mac Studio with M5 Max and M5 Ultra — the ultimate desktop for on‑device AI and the most extreme pro workflows. Apple Newsroom, 25. August 2026. https://www.apple.com/newsroom/2026/08/apple-introduces-new-mac-studio-with-m5-max-and-m5-ultra/ https://www.apple.com/mac-studio/specs/ https://www.apple.com/de/shop/buy-mac/mac-studio (Herstellerangabe) Abgerufen am 2. November 2026.
Arbuthnot, John (1710): An Argument for Divine Providence, Taken from the Constant Regularity Observ’d in the Births of Both Sexes. Philosophical Transactions of the Royal Society of London 27, S. 186–190, DOI 10.1098/rstl.1710.0011. https://doi.org/10.1098/rstl.1710.0011 https://archive.org/details/philtrans02909166 Abgerufen am 25. Oktober 2026.
ARC Prize (2026a): Announcing ARC-AGI-3. Blogbeitrag, 25. März 2026. https://arcprize.org/blog/arc-agi-3-launch
ARC Prize (2026b): ARC-AGI-3: A New Challenge for Frontier Agentic Intelligence. Technischer Bericht (PDF, 23 S.), 22. April 2026. https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf
ARC Prize (2026c): OpenAI’s GPT-6 Astra on ARC-AGI-3. Blogbeitrag, 3. September 2026. https://arcprize.org/blog/astra
Arditi, Andy; Obeso, Oscar; Syed, Aaquib u. a. (2024): Refusal in Language Models Is Mediated by a Single Direction. NeurIPS 2024; Preprint arXiv:2406.11717 (v1 vom 17.6.2024, zitiert nach v3), 17. Juni 2024. https://arxiv.org/abs/2406.11717 Abgerufen am 25. Oktober 2026.
Arena (2024): Does Style Matter? Disentangling Style and Substance in Chatbot Arena. Blogbeitrag, 29. August 2024. https://arena.ai/blog/style-control
Arena (2026a): Leaderboard Dataset (lmarena-ai/leaderboard-dataset). Hugging Face, Datensatz mit Beschreibung, laufend aktualisiert. Stand 15., 22. und 28. Oktober 2026. https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset Abgerufen am 2. November 2026.
Arena (2026b): Text Leaderboard. Website. Stand 27. Oktober 2026. https://arena.ai/leaderboard/text Abgerufen am 2. November 2026.
Arena (2026c): LMArena is now Arena. Blogbeitrag, 28. Januar 2026. https://arena.ai/blog/lmarena-is-now-arena
Arena (o. J. a): Arena’s Ranking Method. Arena, Blog (ohne Datum; früher news.lmarena.ai/ranking-method). https://arena.ai/blog/ranking-method Abgerufen am 25. Oktober 2026.
Arena (o. J. b): FAQ. Website. https://arena.ai/faq
Arena (o. J. c): Leaderboard Changelog. Blog, laufend aktualisiert. https://arena.ai/blog/leaderboard-changelog/ Abgerufen am 20. Oktober 2026.
Arena (o. J. d): Statistical Extensions of the Bradley-Terry and Elo Models. Arena, Blog (ohne Datum). https://arena.ai/blog/extended-arena/ Abgerufen am 25. Oktober 2026.
Arntz, Melanie; Dinelli, Michele; Schlenker, Oliver (2026): KI als Jobkiller für Berufseinsteiger? Entwarnung mit Vorbehalt. IAB-Forum, Nürnberg, 13. Oktober 2026. https://iab-forum.de/ki-als-jobkiller-fuer-berufseinsteiger-entwarnung-mit-vorbehalt/ Abgerufen am 19. Oktober 2026.
Artificial Analysis (2026a): AA-Omniscience v1.1: Knowledge and Hallucination Benchmark. Rangliste mit Methodik (Jackson u. a., arXiv:2511.13029). Stand 2. November 2026. https://artificialanalysis.ai/evaluations/omniscience https://arxiv.org/abs/2511.13029 Abgerufen am 2. November 2026.
Artificial Analysis (2026b): Announcing Artificial Analysis Intelligence Index v4.2. Artificial Analysis, Artikel, 4. September 2026. https://artificialanalysis.ai/articles/artificial-analysis-intelligence-index-v4-2 Abgerufen am 20. Oktober 2026.
Artificial Analysis (o. J.): Intelligence Benchmarking Methodology. Website mit Versionshistorie. https://artificialanalysis.ai/methodology/intelligence-benchmarking Abgerufen am 2. November 2026.
arXiv (o. J.): Content Moderation / About arXiv / Endorsement. info.arxiv.org. https://info.arxiv.org/help/moderation/index.html https://info.arxiv.org/about/index.html https://info.arxiv.org/help/endorsement.html Abgerufen am 2. November 2026.
Aubrey, Kyle; Stam, Nick (2025): Inside NVIDIA Blackwell Ultra: The Chip Powering the AI Factory Era. NVIDIA Technical Blog, zuletzt geändert am 23.1.2026, 22. August 2025. https://developer.nvidia.com/blog/inside-nvidia-blackwell-ultra-the-chip-powering-the-ai-factory-era/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Bacaër, Nicolas (2011): Verhulst and the logistic equation (1838), in: A Short History of Mathematical Population Dynamics, Kap. 6, S. 35–39. Springer, London; DOI 10.1007/978-0-85729-115-8_6. https://doi.org/10.1007/978-0-85729-115-8_6 (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Bai, Yuntao u. a. (2022): Constitutional AI: Harmlessness from AI Feedback. Preprint arXiv:2212.08073, Anthropic (über eigene Modelle), 15. Dezember 2022. https://arxiv.org/abs/2212.08073 Abgerufen am 20. Oktober 2026.
Baker, Bowen; Huizinga, Joost; Gao, Leo u. a. (2025): Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation. Preprint arXiv:2503.11926 (OpenAI, über eigene Modelle), 14. März 2025. https://arxiv.org/abs/2503.11926 Abgerufen am 20. Oktober 2026.
Baptista, Eduardo (2025): China’s DeepSeek says its hit AI model cost just $294,000 to train. Reuters, gelesen bei BNN Bloomberg, 18. September 2025. https://www.bnnbloomberg.ca/business/artificial-intelligence/2025/09/18/chinas-deepseek-says-its-hit-ai-model-cost-just-us294000-to-train/ (Pressebericht) Abgerufen am 25. Oktober 2026.
Barres, Victor; Dong, Honghua; Ray, Soham u. a. (2025): τ²-Bench: Evaluating Conversational Agents in a Dual-Control Environment. Preprint, arXiv:2506.07982, 9. Juni 2025. https://arxiv.org/abs/2506.07982
bartowski (2024a): Meta-Llama-3.1-8B-Instruct-GGUF. Hugging Face, Repository angelegt am 23.7.2024, 23. Juli 2024. https://huggingface.co/bartowski/Meta-Llama-3.1-8B-Instruct-GGUF (Sekundärquelle) Abgerufen am 19. Oktober 2026.
bartowski (2024b): Llama-3.3-70B-Instruct-GGUF. Hugging Face, 6. Dezember 2024. https://huggingface.co/bartowski/Llama-3.3-70B-Instruct-GGUF (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Bartz v. Anthropic Settlement Administrator (2026): Anthropic Copyright Settlement: Long-Form Notice und Stand des Verfahrens. Website des gerichtlich genehmigten Vergleichs, 20. Juli 2026. https://www.anthropiccopyrightsettlement.com/ https://www.anthropiccopyrightsettlement.com/documents Abgerufen am 20. Oktober 2026.
Bauduin, Sven (2026): RAM-Preise Mai 2026: Wo gibt es „günstig“ DDR5 und DDR4? PC Games Hardware, 12. Mai 2026. https://www.pcgameshardware.de/RAM-Hardware-154108/Specials/Speicher-Preise-im-Mai-2026-1534291/ (Pressebericht) Abgerufen am 19. Oktober 2026.
BDEW (2025): Erneuerbare Energien und Stromverbrauch 2024/2025. Foliensatz (Bruttostromverbrauch 2025 vorläufig), 8. Dezember 2025. https://www.bdew.de/media/documents/EE_Jahr_2024_2025_2025_12_08_3.pdf Abgerufen am 19. Oktober 2026.
Bean, Andrew M.; Kearns, Ryan Othniel; Romanou, Angelika u. a. (2025): Measuring what Matters: Construct Validity in Large Language Model Benchmarks. NeurIPS 2025, Datasets and Benchmarks Track; Preprint arXiv:2511.04703 (3.11.2025), 42 Autorinnen und Autoren, Letztautor Adam Mahdi (University of Oxford), 3. November 2025. https://arxiv.org/abs/2511.04703 Abgerufen am 25. Oktober 2026.
Becker, Joel; Rush, Nate; Barnes, Elizabeth u. a. (2025): Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity. Preprint arXiv:2507.09089 (METR), 12. Juli 2025. https://arxiv.org/abs/2507.09089 Abgerufen am 25. Oktober 2026.
★ Bender, Emily M.; Gebru, Timnit; McMillan-Major, Angelina u. a. (2021): On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? FAccT ’21, S. 610–623. https://doi.org/10.1145/3442188.3445922 Abgerufen am 19. Oktober 2026.
Bengio, Yoshua u. a. (2026): International AI Safety Report 2026. International AI Safety Report (Vorsitz Yoshua Bengio), 3. Februar 2026. https://internationalaisafetyreport.org/publication/international-ai-safety-report-2026 Abgerufen am 19. Oktober 2026.
Berliner Beauftragte für Datenschutz und Informationsfreiheit (2025): Berliner Datenschutzbeauftragte meldet KI-App DeepSeek in Deutschland bei Apple und Google als rechtswidrigen Inhalt. Pressemitteilung, Berlin, 27. Juni 2025. https://www.datenschutz-berlin.de/pressemitteilung/berliner-datenschutzbeauftragte-meldet-ki-app-deepseek-in-deutschland-bei-apple-und-google-als-rechtswidrigen-inhalt/ Abgerufen am 19. Oktober 2026.
Berliner Beauftragte für Datenschutz und Informationsfreiheit (2026): Datenschutzbeauftragte sensibilisiert für personalisierte Werbefunktion im Chatbot ChatGPT. Pressemitteilung, Berlin, 2. Oktober 2026. https://www.datenschutz-berlin.de/fileadmin/user_upload/pdf/pressemitteilungen/2026/20261002-BlnBDI-PM-Werbung-ChatGPT.pdf Abgerufen am 19. Oktober 2026.
Bertrand, Marianne; Mullainathan, Sendhil (2004): Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination. American Economic Review 94 (4), S. 991–1013; zitiert nach NBER Working Paper 9873 (Juli 2003). https://www.nber.org/papers/w9873 https://doi.org/10.1257/0002828042002561 Abgerufen am 25. Oktober 2026.
Beygelzimer, Alina; Dauphin, Yann N.; Liang, Percy u. a. (2023): Has the Machine Learning Review Process Become More Arbitrary as the Field Has Grown? The NeurIPS 2021 Consistency Experiment. Preprint arXiv:2306.03262, 5. Juni 2023. https://arxiv.org/abs/2306.03262 Abgerufen am 25. Oktober 2026.
Biderman, Dan u. a. (2024b): LoRA Learns Less and Forgets Less. Transactions on Machine Learning Research (8/2024); Preprint arXiv:2405.09673 (Columbia University, Databricks Mosaic Research). https://arxiv.org/abs/2405.09673 Abgerufen am 25. Oktober 2026.
Biotechnology Innovation Organization (o. J.): Jakob Uszkoreit. Sprecherprofil, BIO International Convention (Selbstauskunft). https://www.bio.org/events/bio-international-convention/speakers/1953322 (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Bitkom (2025): Rechenzentren in Deutschland: KI treibt das Wachstum. Presseinformation, Berlin (Studie des Borderstep Instituts), 10. November 2025. https://www.bitkom.org/Presse/Presseinformation/Rechenzentren-Deutschland-KI-treibt-Wachstum (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Bitkom (2026a): Ein Drittel nutzt KI mindestens einmal pro Woche – von Code bis Küche. Presseinformation, Berlin (Befragung durch Bitkom Research, n = 1.003, ab 16 Jahren, KW 8–11/2026), 28. April 2026. https://www.bitkom.org/Presse/Presseinformation/Ein-Drittel-nutzt-KI-mindestens-einmal-pro-Woche Abgerufen am 19. Oktober 2026.
Bitkom (2026b): Erstmals nutzt die Mehrheit der Unternehmen KI. Presseinformation, Berlin (n = 603 Unternehmen ab 20 Beschäftigten, KW 28–33/2026), 14. September 2026. https://www.bitkom.org/Presse/Presseinformation/Erstmals-nutzt-Mehrheit-Unternehmen-KI Abgerufen am 19. Oktober 2026.
Black Forest Labs (2024): Announcing Black Forest Labs. Blogbeitrag; dazu Impressum (BFL GmbH, Freiburg im Breisgau), 1. August 2024. https://bfl.ai/blog/24-08-01-bfl https://bfl.ai/legal/imprint (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Black Forest Labs (2025): FLUX.2: Frontier Visual Intelligence. Blogbeitrag, 25. November 2025. https://bfl.ai/blog/flux-2 (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Black Forest Labs (2026): FLUX 3 Image Klein. Blogbeitrag, 15. Oktober 2026. https://bfl.ai/blog/flux-3-image-klein (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Boboris, Kat (2025): Attention Authors: Updated Practice for Review Articles and Position Papers in arXiv CS Category. arXiv Blog, 31. Oktober 2025. https://blog.arxiv.org/2025/10/31/attention-authors-updated-practice-for-review-articles-and-position-papers-in-arxiv-cs-category/ Abgerufen am 25. Oktober 2026.
Boboris, Kat (2026): Fair Moderation, Equitable Access, and AI: arXiv’s Updated Rate Limit Policy. arXiv Blog (zuletzt geändert 21.10.2026 laut Seitenmetadaten), 1. Oktober 2026. https://blog.arxiv.org/2026/10/01/updated-rate-limit-policy/ Abgerufen am 2. November 2026.
Bogmans, Christian; Gomez-Gonzalez, Patricia; Melina, Giovanni u. a. (2025): AI Needs More Abundant Power Supplies to Keep Driving Economic Growth. IMF Blog (Internationaler Währungsfonds), 13. Mai 2025. https://www.imf.org/en/Blogs/Articles/2025/05/13/ai-needs-more-abundant-power-supplies-to-keep-driving-economic-growth (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Bowyer, Sam; Aitchison, Laurence; Ivanova, Desi R. (2025): Position: Don’t Use the CLT in LLM Evals With Fewer Than a Few Hundred Datapoints. ICML 2025 (PMLR 267). https://arxiv.org/abs/2503.01747 https://proceedings.mlr.press/v267/bowyer25a.html Abgerufen am 19. Oktober 2026.
Bozzo, Alberto (2026): ChatGPT, il Tribunale di Roma annulla la sanzione da 15 milioni del Garante: decide l’autorità capofila. NT+ Diritto (Il Sole 24 Ore), Mailand, 3. Juni 2026. https://ntplusdiritto.ilsole24ore.com/art/chatgpt-tribunale-roma-annulla-sanzione-15-milioni-garante-decide-l-autorita-capofila-AI88UIRD (Pressebericht) Abgerufen am 19. Oktober 2026.
Bradley, Ralph Allan; Terry, Milton E. (1952): Rank Analysis of Incomplete Block Designs: I. The Method of Paired Comparisons. Biometrika 39 (3/4), S. 324–345, Dezember 1952. https://doi.org/10.2307/2334029 Abgerufen am 25. Oktober 2026.
Bricken, Trenton u. a. (2023): Towards Monosemanticity: Decomposing Language Models With Dictionary Learning. Transformer Circuits Thread (Anthropic), nicht begutachtet, 4. Oktober 2023. https://transformer-circuits.pub/2023/monosemantic-features (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Brohan, Anthony; Brown, Noah; Carbajal, Justice u. a. (2023): RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control. Preprint arXiv:2307.15818 (Google DeepMind; Autoren alphabetisch), 28. Juli 2023. https://arxiv.org/abs/2307.15818 Abgerufen am 25. Oktober 2026.
★ Brown, Tom B. u. a. (2020): Language Models are Few-Shot Learners. NeurIPS 2020. https://arxiv.org/abs/2005.14165 Abgerufen am 19. Oktober 2026.
Brynjolfsson, Erik; Chandar, Bharat; Chen, Ruyu (2026): No Widespread Displacement, but the AI Employment Gap for Young Workers Has Widened to 19%. Stanford Digital Economy Lab; Zusammenfassung der überarbeiteten Fassung von „Canaries in the Coal Mine? Six Facts About the Recent Employment Effects of Artificial Intelligence“ (Working Paper, nicht begutachtet), 12. August 2026. https://digitaleconomy.stanford.edu/news/canariesaug26/ Abgerufen am 19. Oktober 2026.
BTU Cottbus-Senftenberg (o. J.): Künstliche Intelligenz Technologie (M.Sc.). Studiengangsseite. https://www.b-tu.de/kuenstliche-intelligenz-technologie-ms Abgerufen am 25. Oktober 2026.
Bundesamt für Sicherheit in der Informationstechnik (2025a): Evasion Attacks on LLMs – Countermeasures in Practice. A Guide to face Prompt Injections, Jailbreaks and Adversarial Attacks. Bonn, Version 1.0 (Pressemitteilung vom 10. November 2025), 6. November 2025. https://www.bsi.bund.de/SharedDocs/Downloads/EN/BSI/KI/Evasion_Attacks_on_LLMs-Countermeasures.pdf?__blob=publicationFile https://www.bsi.bund.de/DE/Service-Navi/Presse/Alle-Meldungen-News/Meldungen/Evasion-Attacks-LLM_251110.html Abgerufen am 19. Oktober 2026.
Bundesamt für Sicherheit in der Informationstechnik (2025b): Die Lage der IT-Sicherheit in Deutschland 2025 (Zusammenfassung). BSI, Bonn, 11. November 2025. https://www.bsi.bund.de/SharedDocs/Downloads/DE/BSI/Publikationen/Lageberichte/Lagebericht2025_Achtseiter.html Abgerufen am 25. Oktober 2026.
Bundesamt für Sicherheit in der Informationstechnik; Polizeiliche Kriminalprävention der Länder und des Bundes (2026): Cybersicherheitsmonitor 2026, mit Fokusthema Online-Betrug & Künstliche Intelligenz. BSI, Website (Digitaler Verbraucherschutz). https://www.bsi.bund.de/DE/Themen/Verbraucherinnen-und-Verbraucher/Leistungen-und-Kooperationen/Digitaler-Verbraucherschutz/Digitalbarometer/digitalbarometer_node.html Abgerufen am 25. Oktober 2026.
Bundesgerichtshof (2026): Verkündungstermin am 17. Dezember 2026 um 8:45 Uhr in Sachen I ZR 281/25 (Erstellen eines Datensatzes für KI-Training) (Verhandlung: 3.9.2026). Terminhinweis, Karlsruhe. https://www.bundesgerichtshof.de/SharedDocs/Termine/DE/Termine/IZR281-25.html Abgerufen am 2. November 2026.
Bundesnetzagentur (2025a): Genehmigung des Szenariorahmens 2025–2037/2045. Bonn, April 2025. https://www.netzentwicklungsplan.de/sites/default/files/2025-04/Genehmigung%20Szenariorahmen%202025_0.pdf Abgerufen am 19. Oktober 2026.
Bundesnetzagentur (2025b): KI-Service Desk der Bundesnetzagentur gestartet. Pressemitteilung, Bonn; dazu Themenseite „KI“ mit KI-Compliance Kompass, 3. Juli 2025. https://www.bundesnetzagentur.de/SharedDocs/Pressemitteilungen/DE/2025/20250703_KI_ServiceDesk.html https://www.bundesnetzagentur.de/DE/Fachthemen/Digitales/KI/start_ki.html Abgerufen am 19. Oktober 2026.
Bundesrat (2026): Beschluss des Bundesrates: Gesetz zur Modernisierung des Produkthaftungsrechts und zur Änderung weiterer Vorschriften. Drucksache 584/26 (Beschluss), 1069. Sitzung, Berlin, 16. Oktober 2026. https://dserver.bundestag.de/brd/2026/0584-26B.pdf Abgerufen am 2. November 2026.
Bundesrechtsanwaltskammer (2026): KI-Halluzination: KG ermahnt Anwältin wegen erfundener Urteile im Schriftsatz. BRAK-Newsroom, Berlin, 8. Juni 2026. https://www.brak.de/newsroom/news/ki-halluzination-kg-ermahnt-anwaeltin-wegen-erfundener-urteile-im-schriftsatz/ (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Bundesrepublik Deutschland (2023): Gesetz zur Steigerung der Energieeffizienz in Deutschland (Energieeffizienzgesetz – EnEfG), § 3 und § 11. gesetze-im-internet.de, Stand 19.10.2026, 13. November 2023. https://www.gesetze-im-internet.de/enefg/__3.html https://www.gesetze-im-internet.de/enefg/__11.html Abgerufen am 20. Oktober 2026.
Bundesrepublik Deutschland (2026a): Gesetz zur Marktüberwachung und Innovationsförderung von künstlicher Intelligenz (KI-Marktüberwachungs-und-Innovationsförderungs-Gesetz – KI-MIG). BGBl. 2026 I Nr. 223, 22. Juli 2026. https://www.gesetze-im-internet.de/ki-mig/BJNR0DF0B0026.html Abgerufen am 19. Oktober 2026.
Bundesrepublik Deutschland (2026b): Gesetz zur Modernisierung des Produkthaftungsrechts und zur Änderung weiterer Vorschriften vom 22. Oktober 2026. BGBl. 2026 I Nr. 321, 27. Oktober 2026. https://www.recht.bund.de/eli/bund/bgbl-1/2026/321 Abgerufen am 2. November 2026.
Bundesrepublik Deutschland (o. J. a): Strafgesetzbuch (StGB), § 203 Verletzung von Privatgeheimnissen. gesetze-im-internet.de, geltende Fassung. https://www.gesetze-im-internet.de/stgb/__203.html Abgerufen am 19. Oktober 2026.
Bundesrepublik Deutschland (o. J. b): Urheberrechtsgesetz (UrhG), § 44b Text und Data Mining; § 60d Text und Data Mining für Zwecke der wissenschaftlichen Forschung. gesetze-im-internet.de, geltende Fassung. https://www.gesetze-im-internet.de/urhg/__44b.html https://www.gesetze-im-internet.de/urhg/__60d.html Abgerufen am 19. Oktober 2026.
Bundeswettbewerb Künstliche Intelligenz (2026): Bundeswettbewerb Künstliche Intelligenz (BWKI). Website, Tübingen. https://www.bw-ki.de/ Abgerufen am 2. November 2026.
Buolamwini, Joy; Gebru, Timnit (2018): Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification. Proceedings of the 1st Conference on Fairness, Accountability and Transparency, PMLR 81, S. 77–91. https://proceedings.mlr.press/v81/buolamwini18a.html Abgerufen am 19. Oktober 2026.
Bureau International des Poids et Mesures (o. J.): SI prefixes. bipm.org. https://www.bipm.org/en/measurement-units/si-prefixes Abgerufen am 19. Oktober 2026.
Cao, Yinzhi; Yang, Junfeng (2015): Towards Making Systems Forget with Machine Unlearning. 2015 IEEE Symposium on Security and Privacy, San Jose (Columbia University), S. 463–480, DOI 10.1109/SP.2015.35. https://doi.org/10.1109/SP.2015.35 https://www.cs.columbia.edu/~junfeng/papers/unlearning-sp15.pdf Abgerufen am 25. Oktober 2026.
★ Carlini, Nicholas u. a. (2021): Extracting Training Data from Large Language Models. 30th USENIX Security Symposium, S. 2633–2650. https://www.usenix.org/conference/usenixsecurity21/presentation/carlini-extracting Abgerufen am 19. Oktober 2026.
Center for AI Safety (o. J. a): Humanity’s Last Exam: Code und Auswertung mit Bewertermodell. GitHub. https://github.com/centerforaisafety/hle
Center for AI Safety (o. J. b): MMLU (Datensatz cais/mmlu). Hugging Face, MIT-Lizenz. https://huggingface.co/datasets/cais/mmlu
Central Statistics Office Ireland (2026): Data Centres Metered Electricity Consumption 2025. CSO, Statistical Release, 7. Juli 2026. https://www.cso.ie/en/releasesandpublications/ep/p-dcmec/datacentresmeteredelectricityconsumption2025/ Abgerufen am 19. Oktober 2026.
Chae, Songeun u. a. (2026): Self- and Other-Labels Induce Bidirectional Bias in LLM Judges. Findings of EMNLP 2026 (laut arXiv). https://arxiv.org/abs/2608.18091 Abgerufen am 19. Oktober 2026.
Chang, Amy; Conley, Nicholas; Santhanalakshmi Ganesan, Harish u. a. (2025): Death by a Thousand Prompts: Open Model Vulnerability Analysis. Preprint, arXiv:2511.03247 (Cisco), 5. November 2025. https://arxiv.org/abs/2511.03247 Abgerufen am 19. Oktober 2026.
Character.AI (2025): Taking Bold Steps to Keep Teen Users Safe on Character.AI. Character.AI, Blog, 29. Oktober 2025. https://blog.character.ai/u18-chat-announcement (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Charlotin, Damien (2026): AI Hallucination Cases Database. Online-Datenbank, CC BY 4.0 (Stand 2.11.2026). https://www.damiencharlotin.com/hallucinations/ Abgerufen am 2. November 2026.
Chen, Mark u. a. (2021): Evaluating Large Language Models Trained on Code. Preprint, arXiv:2107.03374, 7. Juli 2021. https://arxiv.org/abs/2107.03374 Abgerufen am 20. Oktober 2026.
Chen, Yanda u. a. (2025): Reasoning Models Don’t Always Say What They Think. Preprint arXiv:2505.05410 (15 Autoren, Anthropic, Alignment Science Team), 8. Mai 2025. https://arxiv.org/abs/2505.05410 Abgerufen am 20. Oktober 2026.
Chiang, Wei-Lin u. a. (2024): Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference. ICML 2024, PMLR 235, S. 8359–8388. https://arxiv.org/abs/2403.04132
Chollet, François; Knoop, Mike; Kamradt, Gregory u. a. (2025): ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems. Preprint, 17. Mai 2025. https://arxiv.org/abs/2505.11831
Clopper, C. J.; Pearson, E. S. (1934): The Use of Confidence or Fiducial Limits Illustrated in the Case of the Binomial. Biometrika 26 (4): 404–413. https://doi.org/10.1093/biomet/26.4.404 Abgerufen am 19. Oktober 2026.
Coalition for Content Provenance and Authenticity (o. J.): About C2PA. Website (Projekt der Joint Development Foundation). https://c2pa.org/about/ Abgerufen am 25. Oktober 2026.
CodeCarbon (o. J.): CodeCarbon. Website des Open-Source-Werkzeugs. https://codecarbon.io/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Cohere (2026): Cohere and Aleph Alpha sign agreement to become the first transatlantic sovereign AI solution. Cohere, Blog, 16. September 2026. https://cohere.com/blog/cohere-and-aleph-alpha-sign-agreement (Herstellerangabe) Abgerufen am 2. November 2026.
Collier, Kevin; Cui, Jasmine (2025): OpenAI says DeepSeek may have ‘inapproriately’ [sic] used its data. NBC News, 29. Januar 2025. https://www.nbcnews.com/tech/tech-news/openai-says-deepseek-may-inapproriately-used-data-rcna189872 (Pressebericht) Abgerufen am 20. Oktober 2026.
Competition and Markets Authority (2025): Cloud services market investigation – Summary of final decision. London, 31. Juli 2025. https://assets.publishing.service.gov.uk/media/688b20e6ff8c05468cb7b120/summary_of_final_decision.pdf Abgerufen am 19. Oktober 2026.
CompVis (2022): Stable Diffusion v1-4 Model Card. Hugging Face. https://huggingface.co/CompVis/stable-diffusion-v1-4 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Conley, Nicholas; Chang, Amy (2026): Proprietary Problems: No Frontier Model Is Multi-Turn Immune. Cisco Blogs, 27. Mai 2026. https://blogs.cisco.com/ai/proprietary-problems Abgerufen am 19. Oktober 2026.
Cooper, A. Feder; Choquette-Choo, Christopher A.; Bogen, Miranda u. a. (2024): Machine Unlearning Doesn’t Do What You Think: Lessons for Generative AI Policy and Research. NeurIPS 2025 (Oral); Preprint arXiv:2412.06966 (v1 9.12.2024, v2 31.10.2025), 37 Autorinnen und Autoren, 9. Dezember 2024. https://arxiv.org/abs/2412.06966 Abgerufen am 25. Oktober 2026.
Cooper, A. Feder; Lemley, Mark A.; Casasola, Allison u. a. (2026): Extracting memorized pieces of (copyrighted) books from open-weight language models. Preprint, arXiv:2505.12546 (v1 18.5.2025, v6 20.7.2026; COLM 2026), 20. Juli 2026. https://arxiv.org/abs/2505.12546 Abgerufen am 19. Oktober 2026.
CoreWeave (2026): Pricing (GPU-Instanzen, On-Demand). coreweave.com. https://www.coreweave.com/pricing (Herstellerangabe) Abgerufen am 2. November 2026.
Costello, Thomas H.; Pennycook, Gordon; Rand, David G. (2024): Durably reducing conspiracy beliefs through dialogues with AI. Science 385 (6714), eadq1814, 13. September 2024. https://doi.org/10.1126/science.adq1814 Abgerufen am 25. Oktober 2026.
Cottier, Ben; Snodin, Ben; Owen, David u. a. (2025): LLM inference prices have fallen rapidly but unequally across tasks. Epoch AI, Data Insights, 12. März 2025. https://epoch.ai/data-insights/llm-inference-price-trends (Sekundärquelle) Abgerufen am 26. Oktober 2026.
Cuthbertson, Anthony (2023): ChatGPT boss says he’s achieved human-level AI, then says he’s „just memeing“. The Independent, Tech, 27. September 2023. https://www.the-independent.com/tech/chatgpt-ai-agi-sam-altman-openai-b2419449.html (Pressebericht) Abgerufen am 25. Oktober 2026.
Dastin, Jeffrey (2018): Amazon scraps secret AI recruiting tool that showed bias against women. Reuters, San Francisco, 10. Oktober 2018. https://www.reuters.com/article/world/insight-amazon-scraps-secret-ai-recruiting-tool-that-showed-bias-against-women-idUSKCN1MK0AG/ (Pressebericht) Abgerufen am 25. Oktober 2026.
Dastin, Jeffrey; Nellis, Stephen (2023): Focus: For tech giants, AI like Bing and Bard poses billion-dollar search problem. Reuters, 22. Februar 2023. https://www.reuters.com/technology/tech-giants-ai-like-bing-bard-poses-billion-dollar-search-problem-2023-02-22/ (Pressebericht) Abgerufen am 20. Oktober 2026.
Dathathri, Sumanth u. a. (2024): Scalable watermarking for identifying large language model outputs. Nature 634, S. 818–823 (Google DeepMind), DOI 10.1038/s41586-024-08025-4, 23. Oktober 2024. https://doi.org/10.1038/s41586-024-08025-4 Abgerufen am 25. Oktober 2026.
de Vries, Alex (2023): The growing energy footprint of artificial intelligence. Joule 7 (10), 2191–2194 (Commentary); DOI 10.1016/j.joule.2023.09.004, 10. Oktober 2023. https://doi.org/10.1016/j.joule.2023.09.004 Abgerufen am 20. Oktober 2026.
DeepSeek (2025): DeepSeek-R1 Release. DeepSeek API Docs, News, 20. Januar 2025. https://api-docs.deepseek.com/news/news250120 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
DeepSeek (2026): Thinking Mode. DeepSeek API Docs (ohne Datum). https://api-docs.deepseek.com/guides/thinking_mode (Herstellerangabe) Abgerufen am 20. Oktober 2026.
DeepSeek-AI (2024): DeepSeek-V3 Technical Report. Preprint, arXiv:2412.19437, 27. Dezember 2024. https://arxiv.org/abs/2412.19437 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
DeepSeek-AI (2025a): DeepSeek-R1 (Modellkarte). Hugging Face. https://huggingface.co/deepseek-ai/DeepSeek-R1 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
DeepSeek-AI (2025b): Day 6: One More Thing, DeepSeek-V3/R1 Inference System Overview. GitHub, Repository deepseek-ai/open-infra-index (Open-Source Week, Februar 2025), 1. März 2025. https://github.com/deepseek-ai/open-infra-index/blob/main/202502OpenSourceWeek/day_6_one_more_thing_deepseekV3R1_inference_system_overview.md (Herstellerangabe) Abgerufen am 26. Oktober 2026.
Défossez, Alexandre; Mazaré, Laurent; Orsini, Manu u. a. (2024): Moshi: a speech-text foundation model for real-time dialogue. Preprint arXiv:2410.00037 (Kyutai, Paris; v2 2.10.2024), 17. September 2024. https://arxiv.org/abs/2410.00037 Abgerufen am 25. Oktober 2026.
Deglmann, Florian (2025): KI-Stimme von Tochter? Bei Schockanruf 45.000 Euro gefordert. BR24 (mit Informationen von dpa), 6. September 2025. https://www.br.de/nachrichten/bayern/ki-stimme-von-tochter-bei-schockanruf-45-000-euro-gefordert,Uw0fQYS (Pressebericht) Abgerufen am 25. Oktober 2026.
Der Hamburgische Beauftragte für Datenschutz und Informationsfreiheit (2025): Meta beginnt KI-Training mit persönlichen Daten. FAQ, Hamburg, 27. Mai 2025. https://datenschutz-hamburg.de/news/meta-ki-training-mit-persoenlichen-daten Abgerufen am 19. Oktober 2026.
Der Hamburgische Beauftragte für Datenschutz und Informationsfreiheit (2026): Profiling im Chatbot: ChatGPT läuft in Europa bald mit hyperpersonalisierter Werbung. Meldung, Hamburg, 2. Oktober 2026. https://datenschutz-hamburg.de/news/profilbildung-im-chatbot-chatgpt-laeuft-in-europa-bald-mit-hyperpersonalisierter-werbung Abgerufen am 19. Oktober 2026.
Dettmers, Tim; Pagnoni, Artidoro; Holtzman, Ari u. a. (2023): QLoRA: Efficient Finetuning of Quantized LLMs. NeurIPS 2023; Preprint arXiv:2305.14314 (University of Washington), 23. Mai 2023. https://arxiv.org/abs/2305.14314 Abgerufen am 24. Oktober 2026.
Deutsche Telekom (2025): Für ein souveränes Deutschland: Telekom startet Industrial AI Cloud mit NVIDIA. Medieninformation, Bonn, 4. November 2025. https://www.telekom.com/de/newsroom/aktuelles/medieninformationen/2025/11/fuer-ein-souveraenes-deutschland (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Deutsche Telekom (2026): Deutschlands erste KI-Fabrik für die Industrie geht in München offiziell in Betrieb. Medieninformation, 4. Februar 2026. https://www.telekom.com/de/newsroom/aktuelles/medieninformationen/2026/02/deutschlands-erste-ki-fabrik-fuer-die-industrie (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Deutscher Bundestag (2025): Friedrich Merz mit 325 Stimmen zum Bundeskanzler gewählt. Textarchiv, Berlin, 6. Mai 2025. https://www.bundestag.de/dokumente/textarchiv/2025/kw19-de-kanzlerwahl-1062470 Abgerufen am 19. Oktober 2026.
Deutscher Bundestag (2026a): Beschleunigte Umsetzung der Energieeffizienzrichtlinie beraten. Textarchiv, Berlin, 24. September 2026. https://www.bundestag.de/dokumente/textarchiv/2026/kw39-de-energieeffizienzrichtlinie-1211306 Abgerufen am 2. November 2026.
Deutscher Bundestag (2026b): Bundestag beschließt Modernisierung des Produkthaftungsrechts. Textarchiv (kw41), Berlin, 8. Oktober 2026. https://www.bundestag.de/dokumente/textarchiv/2026/kw41-de-produkthaftungsrecht-1222456 Abgerufen am 2. November 2026.
Deutscher Ethikrat (2023): Mensch und Maschine – Herausforderungen durch Künstliche Intelligenz. Stellungnahme, Berlin, 20.3.2023, 20. März 2023. https://www.ethikrat.org/publikationen/stellungnahmen/mensch-und-maschine/ Abgerufen am 25. Oktober 2026.
Devesa, Diego (2025a): llama : add option to override model tensor buffers (Pull Request Nr. 11397). llama.cpp, GitHub (Nutzername slaren; eröffnet 24.1.2025, übernommen 2.4.2025), 2. April 2025. https://github.com/ggml-org/llama.cpp/pull/11397 Abgerufen am 20. Oktober 2026.
Devesa, Diego (2025b): llama : add –n-cpu-moe option (Pull Request Nr. 15077). llama.cpp, GitHub (Nutzername slaren; eröffnet und übernommen am 4.8.2025), 4. August 2025. https://github.com/ggml-org/llama.cpp/pull/15077 Abgerufen am 20. Oktober 2026.
Devlin, Jacob; Chang, Ming-Wei; Lee, Kenton u. a. (2018): BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. Preprint, arXiv v1 vom 11. Oktober 2018. https://arxiv.org/abs/1810.04805v1 Abgerufen am 19. Oktober 2026.
Die Bundesbeauftragte für den Datenschutz und die Informationsfreiheit (2025): KI in Behörden – Datenschutz von Anfang an mitdenken (Handreichung). BfDI, Bonn, 22. Dezember 2025. https://www.bfdi.bund.de/SharedDocs/Downloads/DE/DokumenteBfDI/Dokumente-allg/2025/Handreichung-KI.pdf?__blob=publicationFile&v=1 Abgerufen am 19. Oktober 2026.
Dodge, Jesse u. a. (2020): Fine-Tuning Pretrained Language Models: Weight Initializations, Data Orders, and Early Stopping. Preprint arXiv:2002.06305, 15. Februar 2020. https://arxiv.org/abs/2002.06305 Abgerufen am 25. Oktober 2026.
Dong, Yihe; Cordonnier, Jean-Baptiste; Loukas, Andreas (2021): Attention is not all you need: pure attention loses rank doubly exponentially with depth. ICML 2021, PMLR 139, S. 2793–2803; Preprint arXiv:2103.03404 (v1 vom 5.3.2021). https://arxiv.org/abs/2103.03404 Abgerufen am 25. Oktober 2026.
dpa-AFX (2026): Zu wenig Strom für weitere Rechenzentren im Rhein-Main-Gebiet. dpa-AFX, über onvista, 18. März 2026. https://www.onvista.de/news/2026/03-18-zu-wenig-strom-fuer-weitere-rechenzentren-im-rhein-main-gebiet-0-10-26490979 (Pressebericht) Abgerufen am 19. Oktober 2026.
Dragan, Anca; Popa, Raluca Ada (2026): Developing Gemini 4 Argon responsibly and safely. Google-Blog (Google DeepMind), 13. Oktober 2026. https://blog.google/innovation-and-ai/technology/safety-security/gemini-4-model-safety-security/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Duong, Tat Dat (o. J.): Tiktokenizer. Web-Werkzeug auf Grundlage von tiktoken. https://tiktokenizer.vercel.app Abgerufen am 19. Oktober 2026.
Eberhard Karls Universität Tübingen (o. J.): Machine Learning (Master). Studiengangsseite. https://uni-tuebingen.de/studium/studienangebot/verzeichnis-der-studiengaenge/detail/course/machine-learning-master/ Abgerufen am 2. November 2026.
Eisemann, Nora; Bunk, Stefan; Mukama, Trasias u. a. (2025): Nationwide real-world implementation of AI for cancer detection in population-based mammography screening. Nature Medicine 31, S. 917–924 (Universität zu Lübeck und Vara, Berlin; finanziert von Vara), 7. Januar 2025. https://www.nature.com/articles/s41591-024-03408-6 Abgerufen am 25. Oktober 2026.
ELLIS (2026): ELLIS PhD & Postdoc Program; Call for Applications 2026. European Laboratory for Learning and Intelligent Systems, Website. https://ellis.eu/research/phd-postdoc Abgerufen am 2. November 2026.
★ Eloundou, Tyna; Manning, Sam; Mishkin, Pamela u. a. (2024): GPTs are GPTs: Labor market impact potential of LLMs. Science 384 (6702), 1306–1308; Preprint 2023. https://doi.org/10.1126/science.adj0998 https://arxiv.org/abs/2303.10130 Abgerufen am 19. Oktober 2026.
Elsworth, Cooper; Huang, Keguo; Patterson, David u. a. (2025): Measuring the environmental impact of delivering AI at Google Scale. Preprint, arXiv:2508.15734v1 (Google), 21. August 2025. https://arxiv.org/abs/2508.15734 (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Epoch AI (o. J.): Benchmarking Hub. Datenbank. Stand 19. Oktober 2026 (Datenpaket); Website-Rangliste GPQA Diamond 26. Oktober 2026. https://epoch.ai/benchmarks Abgerufen am 2. November 2026.
Equal Employment Opportunity Commission u. a. (1978): Uniform Guidelines on Employee Selection Procedures, 29 CFR 1607.4 (D). 43 FR 38295, 38312 (25.8.1978), geändert durch 46 FR 63268 (31.12.1981); Code of Federal Regulations. https://www.law.cornell.edu/cfr/text/29/1607.4 Abgerufen am 25. Oktober 2026.
Esser, Patrick u. a. (2024): Scaling Rectified Flow Transformers for High-Resolution Image Synthesis. ICML 2024, PMLR 235, S. 12606–12633; Preprint arXiv:2403.03206 (Stability AI), 5. März 2024. https://arxiv.org/abs/2403.03206 (Herstellerangabe) Abgerufen am 24. Oktober 2026.
Europäische Kommission (2023): Durchführungsbeschluss (EU) 2023/1795 über die Angemessenheit des Schutzniveaus für personenbezogene Daten nach dem Datenschutzrahmen EU-USA. ABl. L 231 vom 20.9.2023, S. 118, 10. Juli 2023. https://eur-lex.europa.eu/eli/dec_impl/2023/1795/oj/eng Abgerufen am 19. Oktober 2026.
Europäische Kommission (2025a): The General-Purpose AI Code of Practice. Shaping Europe’s digital future (Liste der Unterzeichner laufend aktualisiert), 10. Juli 2025. https://digital-strategy.ec.europa.eu/en/policies/contents-code-gpai Abgerufen am 19. Oktober 2026.
Europäische Kommission (2025b): Guidelines on obligations for General-Purpose AI providers (FAQ). Shaping Europe’s digital future, Brüssel, 11. November 2025. https://digital-strategy.ec.europa.eu/en/faqs/guidelines-obligations-general-purpose-ai-providers Abgerufen am 25. Oktober 2026.
Europäische Kommission (2026a): AI Act (Regulatory framework for AI). Shaping Europe’s digital future, Brüssel (Seitenstand 19.10.2026). https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai Abgerufen am 19. Oktober 2026.
Europäische Kommission (2026b): Commission publishes Code of Practice on marking and labelling AI-generated content. Pressemitteilung IP/26/1328, Brüssel, 10. Juni 2026. https://digital-strategy.ec.europa.eu/en/news/commission-publishes-code-practice-marking-and-labelling-ai-generated-content Abgerufen am 25. Oktober 2026.
Europäische Kommission (o. J.): Adequacy decisions. Brüssel (Seitenstand nicht datiert). https://commission.europa.eu/law/law-topic/data-protection/international-dimension-data-protection/adequacy-decisions_en Abgerufen am 19. Oktober 2026.
Europäische Kommission; Hohe Vertreterin der Union für Außen- und Sicherheitspolitik (2018): Action Plan against Disinformation, JOIN(2018) 36 final. Brüssel; dazu Themenseite „Tackling online disinformation“, 5. Dezember 2018. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML?uri=CELEX%3A52018JC0036 https://digital-strategy.ec.europa.eu/en/policies/online-disinformation Abgerufen am 25. Oktober 2026.
Europäisches Parlament; Rat der Europäischen Union (2016): Verordnung (EU) 2016/679 (Datenschutz-Grundverordnung). ABl. L 119 vom 4.5.2016, 27. April 2016. https://eur-lex.europa.eu/eli/reg/2016/679/oj/deu Abgerufen am 19. Oktober 2026.
Europäisches Parlament; Rat der Europäischen Union (2019): Richtlinie (EU) 2019/790 über das Urheberrecht und die verwandten Schutzrechte im digitalen Binnenmarkt. ABl. L 130 vom 17.5.2019, S. 92, 17. April 2019. https://eur-lex.europa.eu/legal-content/DE/TXT/HTML/?uri=CELEX:32019L0790 Abgerufen am 20. Oktober 2026.
Europäisches Parlament; Rat der Europäischen Union (2024b): Richtlinie (EU) 2024/2853 über die Haftung für fehlerhafte Produkte und zur Aufhebung der Richtlinie 85/374/EWG des Rates. ABl. L vom 18.11.2024, 23. Oktober 2024. https://eur-lex.europa.eu/eli/dir/2024/2853/oj/deu Abgerufen am 19. Oktober 2026.
Europäisches Parlament und Rat der Europäischen Union (2024a): Verordnung (EU) 2024/1689 zur Festlegung harmonisierter Vorschriften für künstliche Intelligenz (KI-Verordnung), geändert durch Verordnung (EU) 2026/1744 vom 8. Juli 2026. ABl. L, 12.7.2024; Änderung ABl. L, 24.7.2026, 13. Juni 2024. https://eur-lex.europa.eu/legal-content/DE/TXT/HTML/?uri=CELEX:32024R1689 https://publications.europa.eu/resource/celex/32024R1689 Abgerufen am 19. Oktober 2026.
European Data Protection Board (2024): Opinion 28/2024 on certain data protection aspects related to the processing of personal data in the context of AI models. Brüssel; dazu Pressemitteilung „EDPB opinion on AI models: GDPR principles support responsible AI“ vom 18.12.2024, 17. Dezember 2024. https://www.edpb.europa.eu/system/files/documents/2024-12/edpb_opinion_202428_ai-models_en.pdf https://www.edpb.europa.eu/news/edpb-opinion-on-ai-models-gdpr-principles-support-responsible-ai_en Abgerufen am 19. Oktober 2026.
Fang, Cathy Mengying u. a. (2025): How AI and Human Behaviors Shape Psychosocial Effects of Chatbot Use: A Longitudinal Randomized Controlled Study. Preprint arXiv:2503.17473v1 (MIT Media Lab und OpenAI); Fassung v2 vom 2.10.2025 mit geändertem Titel und Abstract, 21. März 2025. https://arxiv.org/abs/2503.17473 Abgerufen am 25. Oktober 2026.
fast.ai (o. J.): Practical Deep Learning for Coders. Kursseite (ohne Datum). https://course.fast.ai/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Favaro, Marina; Clark, Jack (2026): When AI builds itself. The Anthropic Institute, Website (Aktualisierung vom 18.9.2026), 4. Juni 2026. https://www.anthropic.com/institute/recursive-self-improvement (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Federal Communications Commission (2024): FCC Fines Man Behind Election Interference Scheme $6 Million for Sending Illegal Robocalls That Used Deepfake Generative AI Technology. Pressemitteilung, Washington (Forfeiture Order FCC 24-104), 26. September 2024. https://docs.fcc.gov/public/attachments/DOC-405811A1.pdf Abgerufen am 19. Oktober 2026.
Federal Trade Commission (2025): FTC Issues Staff Report on AI Partnerships & Investments Study. Washington, D.C.; Pressemitteilung, 17. Januar 2025. https://www.ftc.gov/news-events/news/press-releases/2025/01/ftc-issues-staff-report-ai-partnerships-investments-study https://web.archive.org/web/20251224181357/https://www.ftc.gov/news-events/news/press-releases/2025/01/ftc-issues-staff-report-ai-partnerships-investments-study Abgerufen am 25. Oktober 2026.
Fernandez, Joaquin (2025): The leading generative AI companies. IoT Analytics, 4. März 2025. https://iot-analytics.com/leading-generative-ai-companies/ (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Ferrari Dacrema, Maurizio; Cremonesi, Paolo; Jannach, Dietmar (2019): Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approaches. RecSys 2019. https://arxiv.org/abs/1907.06902 Abgerufen am 19. Oktober 2026.
Field, Hayden (2025): Anthropic will start training its AI models on chat transcripts. The Verge, 28. August 2025. https://www.theverge.com/anthropic/767507/anthropic-user-data-consumers-ai-models-training-privacy (Pressebericht) Abgerufen am 19. Oktober 2026.
Fisher, Ronald A. (1925): Statistical Methods for Research Workers. Edinburgh: Oliver and Boyd; Online-Ausgabe in Classics in the History of Psychology (Christopher D. Green, York University), Abschn. 12. https://psychclassics.yorku.ca/Fisher/Methods/chap3.htm Abgerufen am 25. Oktober 2026.
Fisher, Ronald A. (1935): The Design of Experiments. Edinburgh/London: Oliver and Boyd 1935, Kap. II, Abschn. 5–7; abgeglichen mit der 9. Auflage, New York: Hafner Press 1971. https://archive.org/details/in.ernet.dli.2015.502684 https://archive.org/details/in.ernet.dli.2015.199572 Abgerufen am 2. November 2026.
Forschungszentrum Jülich (2024): Nächster Meilenstein für JUPITER – High-Tech-Zuhause für den europäischen Exascale-Supercomputer. Pressemitteilung, Jülich (Planungsangaben), 23. Januar 2024. https://www.fz-juelich.de/de/aktuelles/news/pressemitteilungen/2024/naechster-meilenstein-fuer-jupiter Abgerufen am 19. Oktober 2026.
Forschungszentrum Jülich (2025): Europas erster Supercomputer erreicht 1 ExaFLOP/s. Pressemitteilung, Jülich, 17. November 2025. https://www.fz-juelich.de/de/aktuelles/news/pressemitteilungen/2025/europas-erster-supercomputer-erreicht-1-exaflop-s Abgerufen am 19. Oktober 2026.
Forschungszentrum Jülich (2026): JUPITER – Die neue Dimension des Rechnens. Website des Forschungszentrums (letzte Änderung 19.6.2026), 19. Juni 2026. https://www.fz-juelich.de/de/jupiter Abgerufen am 19. Oktober 2026.
Fraunhofer IAIS; Fraunhofer IIS (2024): Multilingual und Open Source: Forschungsprojekt OpenGPT-X veröffentlicht großes KI-Sprachmodell. Presseinformation, Sankt Augustin und Erlangen, 26. November 2024. https://www.iais.fraunhofer.de/content/dam/iais/pr/pi/2024/pi_teuken/2024_11_26_Presseinformation_OpenGPT-X-Release_Teuken.pdf (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Friedrich, Felix; Hämmerl, Katharina; Schramowski, Patrick u. a. (2025): Multilingual Text-to-Image Generation Magnifies Gender Stereotypes. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025, Long Papers), S. 19656–19679; Preprint arXiv:2401.16092 (29.1.2024, unter längerem Titel), 29. Januar 2024. https://doi.org/10.18653/v1/2025.acl-long.966 https://arxiv.org/abs/2401.16092 Abgerufen am 25. Oktober 2026.
Fulton, Julian; Norton, Michael; Shilling, Fraser (2019): Water-indexed benefits and impacts of California almonds. Ecological Indicators 96, 711–717; DOI 10.1016/j.ecolind.2017.12.063. https://doi.org/10.1016/j.ecolind.2017.12.063 Abgerufen am 19. Oktober 2026.
FutureHouse (2025): About 30% of Humanity’s Last Exam chemistry/biology answers are likely wrong. Blogbeitrag; Nachtrag vom 16.9.2025, 23. Juli 2025. https://www.futurehouse.org/research/hle-exam
Ganguli, Deep u. a. (2022): Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned. Preprint, arXiv:2209.07858 (Anthropic), 23. August 2022. https://arxiv.org/abs/2209.07858 (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Gao, Leo; Schulman, John; Hilton, Jacob (2023): Scaling Laws for Reward Model Overoptimization. ICML 2023, PMLR 202, S. 10835–10866; Preprint arXiv:2210.10760 (v1 19.10.2022), OpenAI. https://arxiv.org/abs/2210.10760 https://proceedings.mlr.press/v202/gao23h.html Abgerufen am 20. Oktober 2026.
Gao, Leo u. a. (2024): Scaling and evaluating sparse autoencoders. Preprint arXiv:2406.04093 (OpenAI), 6. Juni 2024. https://arxiv.org/abs/2406.04093 (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Garante per la protezione dei dati personali (2023): Provvedimento del 30 marzo 2023 [9870832]. Rom, 30. März 2023. https://www.garanteprivacy.it/home/docweb/-/docweb-display/docweb/9870832 Abgerufen am 19. Oktober 2026.
Garante per la protezione dei dati personali (2025): AI: Il Garante sanziona la società che gestisce il chatbot „Replika“. Pressemitteilung, Rom (doc. web 10132048; Entscheidung vom 10.4.2025), 19. Mai 2025. https://www.garanteprivacy.it/home/docweb/-/docweb-display/docweb/10132048 Abgerufen am 25. Oktober 2026.
Geiping, Jonas; McLeish, Sean; Jain, Neel u. a. (2025): Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach. Preprint arXiv:2502.05171 (7.2.2025); ELLIS Institute Tübingen, MPI für Intelligente Systeme, University of Maryland, Lawrence Livermore National Laboratory, 7. Februar 2025. https://arxiv.org/abs/2502.05171 Abgerufen am 25. Oktober 2026.
GEMA (2026): Gericht entscheidet pro Musikschaffende: GEMA setzt sich gegen SUNO durch und definiert Maßstäbe für internationales Urheberrecht. GEMA News, 31. Juli 2026. https://www.gema.de/de/w/suno-entscheidung-2026 (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Gema, Aryo Pradipta u. a. (2024): Are We Done with MMLU? NAACL 2025 (Long Papers), S. 5069–5096; Preprint 2024. https://arxiv.org/abs/2406.04127
Gerganov, Georgi (2023): Performance of llama.cpp on Apple Silicon M-series. GitHub-Diskussion Nr. 4167, ggml-org/llama.cpp, seit 2023 laufend ergänzt (Community-Messungen). https://github.com/ggml-org/llama.cpp/discussions/4167 (Sekundärquelle) Abgerufen am 20. Oktober 2026.
Gericht der Europäischen Union (2025): Judgment of the General Court in Case T-553/23, Latombe v Commission. Pressemitteilung Nr. 106/25, Luxemburg, 3. September 2025. https://curia.europa.eu/jcms/upload/docs/application/pdf/2025-09/cp250106en.pdf Abgerufen am 19. Oktober 2026.
Germany Trade & Invest (2026): From the Forest to the Stratosphere. GTAI, Markets Germany, 16. Juli 2026. https://www.gtai.de/en/invest/service/publications/markets-germany/from-the-forest-to-the-stratosphere-2011208 (Sekundärquelle) Abgerufen am 20. Oktober 2026.
Geva, Mor; Schuster, Roei; Berant, Jonathan u. a. (2021): Transformer Feed-Forward Layers Are Key-Value Memories. EMNLP 2021; Preprint arXiv:2012.14913 (v1 vom 29.12.2020). https://arxiv.org/abs/2012.14913 Abgerufen am 20. Oktober 2026.
ggml-org (2026a): ggml/src/ggml-metal/kernels/mul_mm.metal und dequantize.h. llama.cpp, GitHub (Metal-Kernel für Apple-Chips, Stand 20.10.2026). https://github.com/ggml-org/llama.cpp/blob/master/ggml/src/ggml-metal/kernels/mul_mm.metal https://github.com/ggml-org/llama.cpp/blob/master/ggml/src/ggml-metal/kernels/dequantize.h Abgerufen am 20. Oktober 2026.
ggml-org (2026b): src/llama-quant.cpp. llama.cpp, GitHub (Stand 19.10.2026). https://github.com/ggml-org/llama.cpp/blob/master/src/llama-quant.cpp Abgerufen am 19. Oktober 2026.
GHG Protocol (2024): IPCC Global Warming Potential Values (Version 2.0). Greenhouse Gas Protocol, 7. August 2024. https://ghgprotocol.org/sites/default/files/2024-08/Global-Warming-Potential-Values%20%28August%202024%29.pdf Abgerufen am 19. Oktober 2026.
Gold, Ashley (2026): Google and Character.AI agree to settle lawsuits over teen suicides. Axios, Technology, 7. Januar 2026. https://www.axios.com/2026/01/07/google-character-ai-lawsuits-teen-suicides (Pressebericht) Abgerufen am 25. Oktober 2026.
Golde, Michael (2016): Rebound-Effekte – Empirische Ergebnisse und Handlungsstrategien. Umweltbundesamt, Hintergrundpapier, Dessau-Roßlau, Juni 2016. https://www.umweltbundesamt.de/system/files/medien/376/publikationen/rebound-effekte_empirische_ergebnisse_und_handlungsstrategien_hintergrundpapier.pdf Abgerufen am 19. Oktober 2026.
Gommers, Jessie; Hernström, Veronica; Josefsson, Viktoria u. a. (2026): Interval cancer, sensitivity, and specificity comparing AI-supported mammography screening with standard double reading without AI in the MASAI study: a randomised, controlled, non-inferiority, single-blinded, population-based, screening-accuracy trial. The Lancet 407 (10527), S. 505–514, DOI 10.1016/S0140-6736(25)02464-X; Letztautorin Kristina Lång (Universität Lund); PubMed 41620232, 31. Januar 2026. https://doi.org/10.1016/S0140-6736(25)02464-X https://pubmed.ncbi.nlm.nih.gov/41620232/ Abgerufen am 25. Oktober 2026.
Good, Irving John (1965): Speculations Concerning the First Ultraintelligent Machine. Advances in Computers 6, S. 31–88, Academic Press, DOI 10.1016/S0065-2458(08)60418-0; Nachdruck Virginia Tech, Technical Report 05-3. https://doi.org/10.1016/S0065-2458(08)60418-0 https://vtechworks.lib.vt.edu/bitstream/handle/10919/89424/TechReport05-3.pdf Abgerufen am 25. Oktober 2026.
Goodfellow, Ian; Bengio, Yoshua; Courville, Aaron (2016a): Deep Learning (Abschnitt Notation). MIT Press. https://www.deeplearningbook.org/contents/notation.html https://github.com/goodfeli/dlbook_notation Abgerufen am 19. Oktober 2026.
Goodfellow, Ian; Bengio, Yoshua; Courville, Aaron (2016b): Deep Learning, Kap. 6 „Deep Feedforward Networks“, Abschn. 6.5 „Back-Propagation and Other Differentiation Algorithms“. MIT Press, Cambridge (Mass.). https://www.deeplearningbook.org/contents/mlp.html Abgerufen am 24. Oktober 2026.
Google (2025a): Fuel your creativity with new generative media models and tools. Google-Blog, 20. Mai 2025. https://blog.google/innovation-and-ai/products/generative-media-models-io-2025/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Google (2025b): Gemini API I/O updates. Google Developers Blog, 23. Mai 2025. https://developers.googleblog.com/en/gemini-api-io-updates/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Google (2025c): Gemini adds Temporary Chats and new personalization features. Google Blog, 13. August 2025. https://blog.google/products-and-platforms/products/gemini/temporary-chats-privacy-controls/ https://support.google.com/gemini/answer/13594961?hl=de (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Google (2026a): Gemini 4 Argon. Gemini-API-Dokumentation („Latest update: October 2026“), inzwischen unter aistudio.google.com. https://aistudio.google.com/docs/models/gemini-4-argon https://ai.google.dev/gemini-api/docs/models/gemini-4-argon (Herstellerangabe) Abgerufen am 2. November 2026.
Google (2026b): Google AI Pro und Ultra – Zugriff auf Gemini 4 Argon und mehr. gemini.google (Preise für Deutschland). https://gemini.google/subscriptions/ (Herstellerangabe) Abgerufen am 2. November 2026.
Google (2026c): Manage & delete your activity in Gemini Apps. Gemini Apps Help. https://support.google.com/gemini/answer/13278892?hl=en (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Google (2026d): Text generation; Gemini 3 developer guide. Gemini-API-Dokumentation. https://aistudio.google.com/docs/text-generation https://ai.google.dev/gemini-api/docs/gemini-3 (Herstellerangabe) Abgerufen am 2. November 2026.
Google (2026e): Use Gemini Apps (Abschnitt „Start a temporary chat“). Gemini Apps Help. https://support.google.com/gemini/answer/13275745?hl=en (Herstellerangabe) Abgerufen am 2. November 2026.
Google (2026f): 2026 Environmental Report. Google Sustainability; Berichtsjahr 2025, 30. Juni 2026. https://sustainability.google/reports/google-2026-environmental-report/ (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Google (2026g): Gemini-Apps-Versionshinweise, Eintrag „Gemini 4 Argon“. gemini.google, 13. Oktober 2026. https://gemini.google/release-notes/ (Herstellerangabe) Abgerufen am 2. November 2026.
Google (2026h): Gemini thinking. Gemini-API-Dokumentation (zuletzt aktualisiert 14.10.2026), 14. Oktober 2026. https://ai.google.dev/gemini-api/docs/thinking (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Google (2026i): Gemini Developer API pricing. Gemini-API-Dokumentation (ai.google.dev leitet seit spätestens 2.11.2026 auf aistudio.google.com/docs/pricing um), 22. Oktober 2026. https://aistudio.google.com/docs/pricing https://ai.google.dev/gemini-api/docs/pricing (Herstellerangabe) Abgerufen am 2. November 2026.
Google (2026j): Gemini Apps Privacy Hub (mit Gemini Apps Privacy Notice). Gemini Apps Help, 26. Oktober 2026. https://support.google.com/gemini/answer/13594961?hl=en (Herstellerangabe) Abgerufen am 2. November 2026.
Google (o. J.): Efficiency. Google Data Centers, laufend aktualisiert. Stand 2. Quartal 2026. https://datacenters.google/efficiency/ (Herstellerangabe) Abgerufen am 2. November 2026.
Google AI Developers Forum (2025): Massive Regression: Detailed Gemini Thinking Process vanished from AI Studio. Forumsfaden ab 20.5.2025; Beiträge #5 von chunduriv (Moderator mit Google-Kennzeichen, 20.5.2025) und #103 von Logan Kilpatrick (Google, 2.6.2025), 20. Mai 2025. https://discuss.ai.google.dev/t/massive-regression-detailed-gemini-thinking-process-vanished-from-ai-studio/83916 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Google DeepMind (2025): AlphaEvolve: A Gemini-powered coding agent for designing advanced algorithms. Google-DeepMind-Blog, 14. Mai 2025. https://deepmind.google/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/ (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Google for Developers (2026): Machine Learning Glossary. Online-Glossar (Stand 10.4.2026), 10. April 2026. https://developers.google.com/machine-learning/glossary (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Google PAIR (o. J.): PAIR Guidebook. Google, People + AI Research. https://pair.withgoogle.com/guidebook/ (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Google Threat Intelligence Group (2025): GTIG AI Threat Tracker: Advances in Threat Actor Usage of AI Tools. Google Cloud Blog, 5. November 2025. https://cloud.google.com/blog/topics/threat-intelligence/threat-actor-usage-of-ai-tools (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Goyal, Priya; Dollár, Piotr; Girshick, Ross u. a. (2017): Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour. Preprint arXiv:1706.02677 (v1 vom 8.6.2017), 8. Juni 2017. https://arxiv.org/abs/1706.02677 Abgerufen am 20. Oktober 2026.
Graham, Paul (2002): A Plan for Spam. Essay, August 2002. https://paulgraham.com/spam.html Abgerufen am 19. Oktober 2026.
Greenblatt, Ryan u. a. (2024): Alignment faking in large language models. Preprint arXiv:2412.14093 (Anthropic und Redwood Research), 18. Dezember 2024. https://arxiv.org/abs/2412.14093 (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Greenhouse Gas Protocol (2026): Scope 2 Market-based Method Update: From 2025-2026 Public Consultation to Current Status. GHG Protocol, Blog, 16. Oktober 2026. https://ghgprotocol.org/blog/scope-2-market-based-method-update-2025-2026-public-consultation-current-status Abgerufen am 2. November 2026.
Greshake, Kai; Abdelnabi, Sahar; Mishra, Shailesh u. a. (2023): Not what you’ve signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection. AISec ’23, Proceedings of the 16th ACM Workshop on Artificial Intelligence and Security, S. 79–90, DOI 10.1145/3605764.3623985; Preprint arXiv:2302.12173 (v1 23.2.2023, v2 5.5.2023). https://doi.org/10.1145/3605764.3623985 https://arxiv.org/abs/2302.12173 Abgerufen am 19. Oktober 2026.
Gridgeman, Norman T.; Mac Lane, Saunders (2008): Wilson, Edwin Bidwell. Complete Dictionary of Scientific Biography, Charles Scribner’s Sons 2008; Abdruck im MacTutor-Archiv (University of St Andrews) und auf Encyclopedia.com. https://mathshistory.st-andrews.ac.uk/DSB/Wilson_Edwin.pdf (Sekundärquelle) Abgerufen am 25. Oktober 2026.
★ Guo, Daya u. a. (2025): DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning. Nature 645, 633–638 (DeepSeek-AI). https://doi.org/10.1038/s41586-025-09422-z https://arxiv.org/abs/2501.12948 Abgerufen am 19. Oktober 2026.
Han, Ziwen; Mankikar, Meher; Michael, Julian u. a. (2025): Search-Time Data Contamination. NeurIPS 2025, Workshop „Evaluating the Evolving LLM Lifecycle“ (Poster); Preprint, 12. August 2025. https://arxiv.org/abs/2508.13180 https://neurips.cc/virtual/2025/122466 Abgerufen am 19. Oktober 2026.
Hao, Shibo; Sukhbaatar, Sainbayar; Su, DiJia u. a. (2024): Training Large Language Models to Reason in a Continuous Latent Space. COLM 2025; Preprint arXiv:2412.06769 (v1 9.12.2024, zitiert nach v4 vom 23.8.2026), FAIR at Meta und UC San Diego, 9. Dezember 2024. https://arxiv.org/abs/2412.06769 Abgerufen am 25. Oktober 2026.
Haviv, Adi; Ram, Ori; Press, Ofir u. a. (2022): Transformer Language Models without Positional Encodings Still Learn Positional Information. Findings of EMNLP 2022; Preprint arXiv:2203.16634 (v1 vom 30.3.2022, v2 vom 5.12.2022), 30. März 2022. https://arxiv.org/abs/2203.16634 Abgerufen am 23. Oktober 2026.
He, Horace; Thinking Machines Lab (2025): Defeating Nondeterminism in LLM Inference. Thinking Machines Lab: Connectionism (Blog), DOI 10.64434/tml.20250910, 10. September 2025. https://thinkingmachines.ai/blog/defeating-nondeterminism-in-llm-inference/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
He, Kaiming; Zhang, Xiangyu; Ren, Shaoqing u. a. (2016): Deep Residual Learning for Image Recognition. CVPR 2016; Preprint arXiv:1512.03385 (v1 vom 10.12.2015). https://arxiv.org/abs/1512.03385 Abgerufen am 20. Oktober 2026.
Heaven, Will Douglas (2026): Mechanistic interpretability: 10 Breakthrough Technologies 2026. MIT Technology Review, 12. Januar 2026. https://www.technologyreview.com/2026/01/12/1130003/mechanistic-interpretability-ai-research-models-2026-breakthrough-technologies/ (Pressebericht) Abgerufen am 25. Oktober 2026.
Heidrich Rechtsanwälte (2026): Pressemeldung Verfahren LAION. Hannover (Prozessvertreter von LAION in den Vorinstanzen), 3. September 2026. https://www.recht-im-internet.de/presseanfragen/pressemeldung-laion (Sekundärquelle) Abgerufen am 20. Oktober 2026.
Heikkilä, Melissa (2023): Making an image with generative AI uses as much energy as charging your phone. MIT Technology Review, 1. Dezember 2023. https://www.technologyreview.com/2023/12/01/1084189/making-an-image-with-generative-ai-uses-as-much-energy-as-charging-your-phone/ (Pressebericht) Abgerufen am 19. Oktober 2026.
★ Hendrycks, Dan u. a. (2021): Measuring Massive Multitask Language Understanding. ICLR 2021; Preprint 2020. https://arxiv.org/abs/2009.03300
Hinton, Geoffrey; Vinyals, Oriol; Dean, Jeff (2015): Distilling the Knowledge in a Neural Network. Preprint, arXiv:1503.02531 (NIPS 2014 Deep Learning Workshop), 9. März 2015. https://arxiv.org/abs/1503.02531 Abgerufen am 20. Oktober 2026.
Ho, Jonathan; Jain, Ajay; Abbeel, Pieter (2020): Denoising Diffusion Probabilistic Models. Preprint arXiv:2006.11239 (UC Berkeley), 19. Juni 2020. https://arxiv.org/abs/2006.11239 Abgerufen am 20. Oktober 2026.
★ Hoffmann, Jordan u. a. (2022): Training Compute-Optimal Large Language Models. NeurIPS 2022. https://arxiv.org/abs/2203.15556 Abgerufen am 19. Oktober 2026.
Holtzman, Ari; Buys, Jan; Du, Li u. a. (2019): The Curious Case of Neural Text Degeneration. ICLR 2020; Preprint arXiv:1904.09751, 22. April 2019. https://arxiv.org/abs/1904.09751 Abgerufen am 20. Oktober 2026.
Hölzle, Urs (2009): Powering a Google search. Google Public Policy Blog, 12. Januar 2009. https://publicpolicy.googleblog.com/2009/01/powering-google-search.html (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Hong, Kelly; Troynikov, Anton; Huber, Jeff (2025): Context Rot: How Increasing Input Tokens Impacts LLM Performance. Chroma, Technical Report (Forschungsbericht eines Vektordatenbank-Herstellers, nicht begutachtet), 14. Juli 2025. https://www.trychroma.com/research/context-rot Abgerufen am 23. Oktober 2026.
Hsieh, Cheng-Ping; Sun, Simeng; Kriman, Samuel u. a. (2024): RULER: What’s the Real Context Size of Your Long-Context Language Models? COLM 2024; Preprint arXiv:2404.06654 (v1 vom 9.4.2024, v3 vom 6.8.2024; NVIDIA), 9. April 2024. https://arxiv.org/abs/2404.06654 Abgerufen am 23. Oktober 2026.
★ Hu, Edward J. u. a. (2022): LoRA: Low-Rank Adaptation of Large Language Models. ICLR 2022; Preprint 2021. https://arxiv.org/abs/2106.09685 Abgerufen am 19. Oktober 2026.
Huang, Wei; Zheng, Xingyu; Ma, Xudong u. a. (2024): How Good Are Low-bit Quantized LLaMA3 Models? An Empirical Study. Preprint, arXiv:2404.14047 (zitiert nach v1; ab v3 vom 13.1.2025 unter dem Titel „An Empirical Study of LLaMA3 Quantization: From LLMs to MLLMs“, die Zahlen aus Tab. 1 dort unverändert in Tab. 3), 22. April 2024. https://arxiv.org/abs/2404.14047v1 Abgerufen am 20. Oktober 2026.
Hubbard, Katharine E.; Dunbar, Sonja D. (2017): Perceptions of scientific research literature and strategies for reading papers depend on academic career stage. PLOS ONE 12 (12): e0189753, 28. Dezember 2017. https://doi.org/10.1371/journal.pone.0189753 Abgerufen am 19. Oktober 2026.
Hugging Face (2023): What’s going on with the Open LLM Leaderboard? Hugging Face Blog, 23. Juni 2023. https://huggingface.co/blog/open-llm-leaderboard-mmlu
Hugging Face (2026a): Modellseiten und Lizenzdateien von DeepSeek-V4-Pro, Qwen3.8-27B, Qwen3.8-2.4T-A95B und Kimi-K3. Hugging Face. https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro https://huggingface.co/Qwen/Qwen3.8-27B https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B https://huggingface.co/moonshotai/Kimi-K3 (Herstellerangabe) Abgerufen am 2. November 2026.
Hugging Face (2026b): Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident. Hugging Face, Blog; dazu „Security incident disclosure — July 2026“ (16.7.2026), 27. Juli 2026. https://huggingface.co/blog/agent-intrusion-technical-timeline https://huggingface.co/blog/security-incident-july-2026 Abgerufen am 25. Oktober 2026.
Hugging Face (o. J. a): Daily Papers (Hugging Face Papers). Website; paperswithcode.com leitet auf huggingface.co/papers/trending weiter. https://huggingface.co/papers https://huggingface.co/papers/trending (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Hugging Face (o. J. b): Hugging Face Learn. Kursübersicht (ohne Datum). https://huggingface.co/learn (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Hugging Face Hub (2026): Model-API: eigene Zählung der Modelle mit „abliterated“ im Namen. Hugging Face Hub, öffentliche Model-API (Dokumentation), 2. November 2026. https://huggingface.co/docs/hub/api Abgerufen am 2. November 2026.
Hugging-Face-Nutzer (Name bewusst nicht genannt) (2026): Modellkarte einer Kolibri-1-Variante mit weitgehend entferntem Ablehnen (Name bewusst nicht genannt). Hugging Face, Repository angelegt am 5.10.2026 (laut API); Ablehnungszahlen Selbstangabe des Hochladenden. https://huggingface.co (Herstellerangabe) Abgerufen am 19. Oktober 2026.
IBM (o. J.): Deep Blue. IBM Heritage. https://www.ibm.com/history/deep-blue (Herstellerangabe) Abgerufen am 19. Oktober 2026.
ICLR (2012): ICLR 2013 Call for Papers (International Conference on Learning Representations). Spiegelung im ACL-Portal, 29. Oktober 2012. https://aclweb.org/portal/node/1419 Abgerufen am 19. Oktober 2026.
ICLR (2025a): Formatting Instructions for ICLR 2026 Conference Submissions (Abschnitt Default Notation). GitHub ICLR/Master-Template, iclr2026_conference.tex. https://github.com/ICLR/Master-Template Abgerufen am 19. Oktober 2026.
ICLR (2025b): ICLR 2026 Author Guide. iclr.cc. https://iclr.cc/Conferences/2026/AuthorGuide Abgerufen am 19. Oktober 2026.
ICLR (2025c): Announcing the Outstanding Paper Awards at ICLR 2025. ICLR Blog, 22. April 2025. https://blog.iclr.cc/2025/04/22/announcing-the-outstanding-paper-awards-at-iclr-2025/ Abgerufen am 19. Oktober 2026.
ICLR 2026 Program Chairs (2026): A Retrospective on the ICLR 2026 Review Process. ICLR Blog, 31. März 2026. https://blog.iclr.cc/2026/03/31/a-retrospective-on-the-iclr-2026-review-process/ Abgerufen am 19. Oktober 2026.
ICML (2026a): ICML 2026 Call for Papers. icml.cc, 25. Januar 2026. https://icml.cc/Conferences/2026/CallForPapers https://icml.cc/Conferences/2026/AuthorInstructions Abgerufen am 19. Oktober 2026.
ICML 2026 Program Chairs (2026): On Violations of LLM Review Policies. ICML Blog (Update 26.3.2026); Program Chairs Agarwal, Dudík, Li, Jaggi, Scientific Integrity Chair Shah, Communications Chairs Gorman, Kamath, 18. März 2026. https://blog.icml.cc/2026/03/18/on-violations-of-llm-review-policies/ Abgerufen am 19. Oktober 2026.
ifo Institut (2025): Ein Viertel der Unternehmen rechnet mit Stellenabbau durch Künstliche Intelligenz. Pressemitteilung (ifo Konjunkturumfrage), München, 5. Juni 2025. https://www.ifo.de/fakten/2025-06-05/ein-viertel-der-unternehmen-rechnet-mit-stellenabbau-durch-kuenstliche Abgerufen am 20. Oktober 2026.
Inan, Hakan u. a. (2023): Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations. Preprint, arXiv:2312.06674 (Meta), 7. Dezember 2023. https://arxiv.org/abs/2312.06674 (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Initiative D21 (2025): D21-Digital-Index 2024/2025. Jährliches Lagebild zur Digitalen Gesellschaft. Berlin; durchgeführt von Kantar, gefördert vom BMWK, 10. März 2025. https://initiatived21.de/publikationen/d21-digital-index/2024-25 https://initiatived21.de/download/248067/D21DigitalIndex_2024-2025.pdf Abgerufen am 19. Oktober 2026.
Initiative Urheberrecht (2025): OpenAI legt Berufung ein. urheber.info, 10. Dezember 2025. https://urheber.info/diskurs/openai-legt-berufung-ein (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Initiative Urheberrecht (2026): Weiterleitung an EuGH wahrscheinlich. urheber.info, Berlin (Verband; stützt sich auf dpa und FAZ), 4. September 2026. https://urheber.info/diskurs/weiterleitung-an-eugh-wahrscheinlich (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Intel (o. J. a): Intel Core i9-14900K Processor – Specifications. Intel ARK. https://www.intel.com/content/www/us/en/products/sku/236773/intel-core-i9-processor-14900k-36m-cache-up-to-6-00-ghz/specifications.html (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Intel (o. J. b): Intel Core Ultra 9 Processor 285K – Specifications. Intel ARK (Produkt eingeführt Q4/2024). https://www.intel.com/content/www/us/en/products/sku/241060/intel-core-ultra-9-processor-285k-36m-cache-up-to-5-70-ghz/specifications.html (Herstellerangabe) Abgerufen am 19. Oktober 2026.
interface (früher Stiftung Neue Verantwortung) (o. J.): interface – Tech analysis and policy ideas for Europe. Website, Berlin. https://www.interface-eu.org/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
International Energy Agency (2024): Electricity 2024 – Analysis and forecast to 2026. IEA, Paris, Januar 2024. https://iea.blob.core.windows.net/assets/18f3ed24-4b26-4c83-a3d2-8a1be51c8cc8/Electricity2024-Analysisandforecastto2026.pdf Abgerufen am 19. Oktober 2026.
International Energy Agency (2025): Energy and AI. IEA, Paris; Bericht (Abschnitte „Energy demand from AI“ und Executive Summary), 10. April 2025. https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai https://www.iea.org/reports/energy-and-ai/executive-summary Abgerufen am 2. November 2026.
International Energy Agency (2026a): Key Questions on Energy and AI. IEA, Paris, 16. April 2026. https://www.iea.org/reports/key-questions-on-energy-and-ai https://iea.blob.core.windows.net/assets/3179f7f8-01f6-4dd6-bffa-c9f7b73f1dc9/KeyQuestionsonEnergyandAI.pdf Abgerufen am 19. Oktober 2026.
International Energy Agency (2026b): Powering Artificial Intelligence: Strategies for integrating data centres into electricity systems. IEA, Paris; Executive summary, Oktober 2026. https://www.iea.org/reports/powering-artificial-intelligence/executive-summary Abgerufen am 19. Oktober 2026.
ISO (2006): ISO 14040:2006 Environmental management – Life cycle assessment – Principles and framework. Norm (Abstract), 2022 bestätigt; Ergänzung Amd 1:2020, Juli 2006. https://committee.iso.org/standard/37456.html Abgerufen am 19. Oktober 2026.
JEDEC (2025): JEDEC® and Industry Leaders Collaborate to Release JESD270-4 HBM4 Standard: Advancing Bandwidth, Efficiency, and Capacity for AI and HPC. Pressemitteilung, Arlington, 16. April 2025. https://www.jedec.org/news/pressreleases/jedec%C2%AE-and-industry-leaders-collaborate-release-jesd270-4-hbm4-standard-advancing Abgerufen am 19. Oktober 2026.
Jevons, William Stanley (1866): The Coal Question, 2. Aufl., Kap. VII „Of the Economy of Fuel“. Macmillan, London (Erstausgabe 1865). https://www.econlib.org/library/YPDBooks/Jevons/jvnCQ7.html Abgerufen am 19. Oktober 2026.
Jiang, Albert Q. u. a. (2024): Mixtral of Experts. Preprint, arXiv:2401.04088, 8. Januar 2024. https://arxiv.org/abs/2401.04088 Abgerufen am 20. Oktober 2026.
Jimenez, Carlos E. u. a. (2024): SWE-bench: Can Language Models Resolve Real-World GitHub Issues? ICLR 2024. https://arxiv.org/abs/2310.06770
★ Jouppi, Norman P. u. a. (2017): In-Datacenter Performance Analysis of a Tensor Processing Unit. ISCA 2017. https://arxiv.org/abs/1704.04760 Abgerufen am 19. Oktober 2026.
Jurafsky, Daniel; Martin, James H. (2026): Speech and Language Processing, 3. Aufl. (Entwurf), Kap. 1 „Introduction“, Kap. 2 „Words and Tokens“, Kap. 3 „N-gram Language Models“, Kap. 4 „Logistic Regression and Text Classification“, Kap. 5 „Embeddings“, Kap. 7 „Transformers and Pretraining“ und Kap. 8 „Post-training“. Stanford University, Entwurf vom 19.8.2026, 19. August 2026. https://web.stanford.edu/~jurafsky/slp3/1.pdf https://web.stanford.edu/~jurafsky/slp3/2.pdf https://web.stanford.edu/~jurafsky/slp3/3.pdf https://web.stanford.edu/~jurafsky/slp3/4.pdf https://web.stanford.edu/~jurafsky/slp3/5.pdf https://web.stanford.edu/~jurafsky/slp3/7.pdf https://web.stanford.edu/~jurafsky/slp3/8.pdf (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Kaas, Leo; Manger, Christian (2012): Ethnic Discrimination in Germany’s Labour Market: A Field Experiment. German Economic Review 13 (1), S. 1–20; zitiert nach IZA Discussion Paper 4741 (2010). https://www.iza.org/publications/dp/4741/ethnic-discrimination-in-germanys-labour-market-a-field-experiment Abgerufen am 25. Oktober 2026.
Kadavath, Saurav u. a. (2022): Language Models (Mostly) Know What They Know. Preprint arXiv:2207.05221, Anthropic, 11. Juli 2022. https://arxiv.org/abs/2207.05221 Abgerufen am 20. Oktober 2026.
★ Kalai, Adam Tauman; Nachum, Ofir; Vempala, Santosh S. u. a. (2026): Evaluating large language models for accuracy incentivizes hallucinations. Nature 653, 1047–1051 (online 22.4.2026); Preprint 2025 „Why Language Models Hallucinate“. https://doi.org/10.1038/s41586-026-10549-w https://arxiv.org/abs/2509.04664 Abgerufen am 19. Oktober 2026.
Kammergericht Berlin (2025): Beschluss 17 WF 144/25 (Von KI erfundene Rechtsprechungszitate in anwaltlichem Schriftsatz). ECLI:DE:KG:2025:1120.17WF144.25.00; Berliner Vorschriften- und Rechtsprechungsdatenbank; Fundstellen u. a. NJW 2026, 1892; IBR 2026, 485; IMR 2026, 421, 20. November 2025. https://gesetze.berlin.de/perma?d=NJRE001640995 https://dejure.org/2025,51025 Abgerufen am 19. Oktober 2026.
★ Kaplan, Jared u. a. (2020): Scaling Laws for Neural Language Models. Preprint. https://arxiv.org/abs/2001.08361 Abgerufen am 19. Oktober 2026.
Karpathy, Andrej (o. J.): Neural Networks: Zero to Hero. Kursseite mit Videos (ohne Datum). https://karpathy.ai/zero-to-hero.html (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Karpukhin, Vladimir u. a. (2020): Dense Passage Retrieval for Open-Domain Question Answering. EMNLP 2020; Preprint arXiv:2004.04906, 10. April 2020. https://arxiv.org/abs/2004.04906 Abgerufen am 20. Oktober 2026.
Kassianik, Paul; Karbasi, Amin (2025): Evaluating Security Risk in DeepSeek and Other Frontier Reasoning Models. Cisco Blogs (Cisco/Robust Intelligence mit der University of Pennsylvania), 31. Januar 2025. https://blogs.cisco.com/security/evaluating-security-risk-in-deepseek-and-other-frontier-reasoning-models Abgerufen am 19. Oktober 2026.
Kawrakow, Iwan (2023): k-quants (Pull Request Nr. 1684). llama.cpp, GitHub (eröffnet 3.6.2023, übernommen 5.6.2023), 5. Juni 2023. https://github.com/ggml-org/llama.cpp/pull/1684 Abgerufen am 19. Oktober 2026.
Keshav, S. (2007): How to Read a Paper. ACM SIGCOMM Computer Communication Review 37 (3), S. 83–84, DOI 10.1145/1273445.1273458, Juli 2007. https://doi.org/10.1145/1273445.1273458 http://ccr.sigcomm.org/online/files/p83-keshavA.pdf Abgerufen am 19. Oktober 2026.
Keshav, S. (2016): How to Read a Paper (Version of February 17, 2016). nicht formal veröffentlicht; Kopie auf der Kursseite Harvard CS 161 (2019), 17. Februar 2016. https://www.read.seas.harvard.edu/cs161/2019/pdf/keshav16how.pdf Abgerufen am 19. Oktober 2026.
KI-Campus (Stifterverband) (o. J.): KI-Campus – Die Lernplattform für Künstliche Intelligenz. Website mit „Über uns“ (ohne Datum). https://ki-campus.org/ https://ki-campus.org/ueber-uns (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Kingma, Diederik P.; Ba, Jimmy (2014): Adam: A Method for Stochastic Optimization. Preprint arXiv:1412.6980 (v1 vom 22.12.2014); Konferenzfassung ICLR 2015, San Diego, 22. Dezember 2014. https://arxiv.org/abs/1412.6980 Abgerufen am 20. Oktober 2026.
Kirchenbauer, John; Geiping, Jonas; Wen, Yuxin u. a. (2023): A Watermark for Large Language Models. ICML 2023, PMLR 202; Preprint arXiv:2301.10226 (University of Maryland; zitiert nach v4), 24. Januar 2023. https://arxiv.org/abs/2301.10226 Abgerufen am 25. Oktober 2026.
Kleinberg, Jon; Mullainathan, Sendhil; Raghavan, Manish (2017): Inherent Trade-Offs in the Fair Determination of Risk Scores. ITCS 2017; Preprint arXiv:1609.05807 (19.9.2016). https://arxiv.org/abs/1609.05807 Abgerufen am 25. Oktober 2026.
Koenecke, Allison; Choi, Anna Seo Gyeong; Mei, Katelyn X. u. a. (2024): Careless Whisper: Speech-to-Text Hallucination Harms. FAccT 2024, DOI 10.1145/3630106.3658996; Preprint arXiv:2402.08021 (v2 3.5.2024), 12. Februar 2024. https://arxiv.org/abs/2402.08021 Abgerufen am 25. Oktober 2026.
Kojima, Takeshi; Gu, Shixiang Shane; Reid, Machel u. a. (2022): Large Language Models are Zero-Shot Reasoners. Preprint arXiv:2205.11916 (University of Tokyo, Google Research), 24. Mai 2022. https://arxiv.org/abs/2205.11916 Abgerufen am 20. Oktober 2026.
Konferenz der unabhängigen Datenschutzaufsichtsbehörden des Bundes und der Länder (2024): Orientierungshilfe Künstliche Intelligenz und Datenschutz, Version 1.0. DSK, 6. Mai 2024. https://www.datenschutzkonferenz-online.de/media/oh/20240506_DSK_Orientierungshilfe_KI_und_Datenschutz.pdf Abgerufen am 19. Oktober 2026.
Korbak, Tomek; Balesni, Mikita u. a. (2025): Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety. Preprint arXiv:2507.11473 (v1 15.7.2025, v2 7.12.2025), 41 Autorinnen und Autoren u. a. von UK AI Security Institute, Apollo Research, OpenAI, Google DeepMind, Anthropic, Meta; Ansichten der Einzelnen, nicht der Institutionen, 15. Juli 2025. https://arxiv.org/abs/2507.11473 Abgerufen am 25. Oktober 2026.
★ Krizhevsky, Alex; Sutskever, Ilya; Hinton, Geoffrey E. (2012): ImageNet Classification with Deep Convolutional Neural Networks. NIPS 25 (2012). https://papers.nips.cc/paper_files/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html Abgerufen am 19. Oktober 2026.
Kullback, Solomon; Leibler, Richard A. (1951): On Information and Sufficiency. Annals of Mathematical Statistics 22 (1), S. 79–86, März 1951. https://doi.org/10.1214/aoms/1177729694 Abgerufen am 25. Oktober 2026.
Kundel, Dominik (2025): OpenAI Harmony Response Format. OpenAI Cookbook, 5. August 2025. https://developers.openai.com/cookbook/articles/openai-harmony (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Kwa, Thomas; West, Ben; Becker, Joel u. a. (2025): Measuring AI Ability to Complete Long Software Tasks. Preprint arXiv:2503.14499 (METR; v1 vom 18.3.2025, zitiert nach v4), 18. März 2025. https://arxiv.org/abs/2503.14499 Abgerufen am 25. Oktober 2026.
Kwon, Woosuk u. a. (2023): Efficient Memory Management for Large Language Model Serving with PagedAttention. SOSP 2023; Preprint arXiv:2309.06180 (9 Autoren), 12. September 2023. https://arxiv.org/abs/2309.06180 Abgerufen am 19. Oktober 2026.
Kyutai (2026): Kyutai: open-science AI lab (Selbstdarstellung). Kyutai, Paris, Website. https://kyutai.org/ (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Lambert, Nathan u. a. (2024): Tülu 3: Pushing Frontiers in Open Language Model Post-Training. Preprint arXiv:2411.15124 (Allen Institute for AI, University of Washington), 22. November 2024. https://arxiv.org/abs/2411.15124 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Landgericht Hamburg (2024): Urteil vom 27. September 2024, 310 O 227/23. Hamburg (Landesrechtsprechung Hamburg), 27. September 2024. https://www.landesrecht-hamburg.de/bsha/document/NJRE001588058 https://openjur.de/u/2495651.html Abgerufen am 20. Oktober 2026.
Landgericht München I (2025): Pressemitteilung 11/2025: Urteil GEMA gegen Open AI (Az. 42 O 14139/24). München, 11. November 2025. https://www.justiz.bayern.de/gerichte-und-behoerden/landgericht/muenchen-1/presse/2025/11.php Abgerufen am 19. Oktober 2026.
Lane, Rupert; Hay, Anthony; Schwarz, Arthur u. a. (2025): ELIZA Reanimated: The world’s first chatbot restored on the world’s first time sharing system. Preprint, arXiv:2501.06707 (v1 12.1.2025); Zeitschriftenfassung „ELIZA Reanimated: Restoring the Mother of All Chatbots to One of the World’s First Time-Sharing Systems“, IEEE Annals of the History of Computing 47 (2), S. 68–76, DOI 10.1109/MAHC.2025.3564095, 12. Januar 2025. https://arxiv.org/abs/2501.06707 Abgerufen am 19. Oktober 2026.
Lapuschkin, Sebastian; Wäldchen, Stephan; Binder, Alexander u. a. (2019): Unmasking Clever Hans predictors and assessing what machines really learn. Nature Communications 10, Artikel 1096, DOI 10.1038/s41467-019-08987-4, 11. März 2019. https://www.nature.com/articles/s41467-019-08987-4 Abgerufen am 19. Oktober 2026.
Lee, Harrison u. a. (2024): RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback. ICML 2024, PMLR 235, S. 26874–26901; Preprint arXiv:2309.00267 (v1 1.9.2023), Google. https://arxiv.org/abs/2309.00267 https://proceedings.mlr.press/v235/lee24t.html Abgerufen am 20. Oktober 2026.
Lee, Katherine u. a. (2022): Deduplicating Training Data Makes Language Models Better. ACL 2022; Preprint arXiv:2107.06499 (v1 14.7.2021). https://arxiv.org/abs/2107.06499 Abgerufen am 19. Oktober 2026.
Lee, Wendy (2024): Scarlett Johansson also thinks OpenAI’s new voice sounds like her. She’s not happy about it. Los Angeles Times, Entertainment & Arts, 20. Mai 2024. https://www.latimes.com/entertainment-arts/business/story/2024-05-20/openai-chatgpt-voice-scarlett-johansson-paused (Pressebericht) Abgerufen am 25. Oktober 2026.
Levy, Steven (2024): 8 Google Employees Invented Modern AI. Here’s the Inside Story. Wired, 20. März 2024. https://www.wired.com/story/eight-google-employees-invented-modern-ai-transformers-paper/ (Pressebericht) Abgerufen am 25. Oktober 2026.
★ Lewis, Patrick u. a. (2020): Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. NeurIPS 2020. https://arxiv.org/abs/2005.11401 Abgerufen am 19. Oktober 2026.
★ Li, Kenneth u. a. (2023a): Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task. ICLR 2023. https://arxiv.org/abs/2210.13382 Abgerufen am 19. Oktober 2026.
Li, Pengfei; Yang, Jianyi; Islam, Mohammad A. u. a. (2023b): Making AI Less “Thirsty”: Uncovering and Addressing the Secret Water Footprint of AI Models. Preprint, Fassungen v1 (6.4.2023) bis v5 (26.3.2025); Communications of the ACM 68 (7), 54–61, 2025. https://arxiv.org/abs/2304.03271 Abgerufen am 19. Oktober 2026.
Liang, Percy; Hashimoto, Tatsunori (2026): CS336: Language Modeling from Scratch (Spring 2026). Kursseite, Stanford. https://cs336.stanford.edu/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Liang, Weixin; Yuksekgonul, Mert; Mao, Yining u. a. (2023a): GPT detectors are biased against non-native English writers. Patterns 4 (7), 100779; Preprint arXiv:2304.02819. https://doi.org/10.1016/j.patter.2023.100779 Abgerufen am 25. Oktober 2026.
Lieberum, Tom; Rajamanoharan, Senthooran; Conmy, Arthur u. a. (2024): Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2. Preprint arXiv:2408.05147 (Google DeepMind), 9. August 2024. https://arxiv.org/abs/2408.05147 (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Lightcap, Brad (2025): How we’re responding to The New York Times’ data demands in order to protect user privacy. OpenAI Blog (Update vom 22.10.2025), 5. Juni 2025. https://openai.com/index/response-to-nyt-data-demands/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Lightman, Hunter u. a. (2023): Let’s Verify Step by Step. Preprint arXiv:2305.20050 (OpenAI), 31. Mai 2023. https://arxiv.org/abs/2305.20050 Abgerufen am 20. Oktober 2026.
Lindsey, Jack u. a. (2025): On the Biology of a Large Language Model. Transformer Circuits Thread (Anthropic), nicht begutachtet, 27. März 2025. https://transformer-circuits.pub/2025/attribution-graphs/biology.html (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Linux Foundation (2025): Linux Foundation Announces the Formation of the Agentic AI Foundation (AAIF), Anchored by New Project Contributions Including Model Context Protocol (MCP), goose and AGENTS.md. Pressemitteilung, San Francisco, 9. Dezember 2025. https://www.linuxfoundation.org/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Lipman, Yaron; Chen, Ricky T. Q.; Ben-Hamu, Heli u. a. (2022): Flow Matching for Generative Modeling. ICLR 2023; Preprint arXiv:2210.02747 (Meta AI FAIR, Weizmann Institute of Science; zitiert nach v2 vom 8.2.2023), 6. Oktober 2022. https://arxiv.org/abs/2210.02747 Abgerufen am 24. Oktober 2026.
Lipton, Zachary C.; Steinhardt, Jacob (2018): Troubling Trends in Machine Learning Scholarship. ICML 2018: The Debates; ACM Queue 17 (1), 2019. https://arxiv.org/abs/1807.03341 Abgerufen am 19. Oktober 2026.
Liu, Haotian; Li, Chunyuan; Wu, Qingyang u. a. (2023): Visual Instruction Tuning. NeurIPS 2023; Preprint arXiv:2304.08485, 17. April 2023. https://arxiv.org/abs/2304.08485 Abgerufen am 20. Oktober 2026.
Liu, Zichen u. a. (2025): Understanding R1-Zero-Like Training: A Critical Perspective. COLM 2025; Preprint arXiv:2503.20783 (Sea AI Lab, National University of Singapore, Singapore Management University). https://arxiv.org/abs/2503.20783 Abgerufen am 25. Oktober 2026.
Llama Team, AI @ Meta (2024): The Llama 3 Herd of Models. Preprint, arXiv:2407.21783v1, 31. Juli 2024. https://arxiv.org/abs/2407.21783 (Herstellerangabe) Abgerufen am 24. Oktober 2026.
LMSYS (o. J.): FastChat, fastchat/serve/monitor/rating_systems.py (Funktion compute_bt). Quellcode der Arena, GitHub. https://github.com/lm-sys/FastChat/blob/main/fastchat/serve/monitor/rating_systems.py Abgerufen am 25. Oktober 2026.
Lüber, Klaus (2026): Zuse Schools: Wie Deutschland KI-Talente anzieht. deutschland.de (veröffentlicht 20.5.2026, geändert 27.5.2026), 20. Mai 2026. https://www.deutschland.de/de/topic/wirtschaft/konrad-zuse-schools-ki-talente-deutschland (Pressebericht) Abgerufen am 25. Oktober 2026.
Luccioni, Alexandra Sasha; Jernite, Yacine; Strubell, Emma (2024): Power Hungry Processing: Watts Driving the Cost of AI Deployment? FAccT 2024, Rio de Janeiro, S. 85–99; DOI 10.1145/3630106.3658542; Vorabfassung arXiv:2311.16863. https://facctconference.org/static/papers24/facct24-6.pdf https://doi.org/10.1145/3630106.3658542 Abgerufen am 19. Oktober 2026.
Luong, Thang; Lockhart, Edward (2025): Advanced version of Gemini with Deep Think officially achieves gold-medal standard at the International Mathematical Olympiad. Google-DeepMind-Blog, 21. Juli 2025. https://deepmind.google/blog/advanced-version-of-gemini-with-deep-think-officially-achieves-gold-medal-standard-at-the-international-mathematical-olympiad/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
MacCallum, Noah; Lee, Julian (2025): GPT-4.1 Prompting Guide. OpenAI Cookbook (archiviert), 14. April 2025. https://developers.openai.com/cookbook/examples/gpt4-1_prompting_guide (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Magesh, Varun; Surani, Faiz; Dahl, Matthew u. a. (2025): Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools. Journal of Empirical Legal Studies 22 (2), S. 216–242, DOI 10.1111/jels.12413; Preprint arXiv:2405.20362 (30.5.2024). https://doi.org/10.1111/jels.12413 https://arxiv.org/abs/2405.20362 Abgerufen am 19. Oktober 2026.
Mathematical Association of America (o. J.): American Invitational Mathematics Examination (AIME). Website der MAA (AMC-Programme). https://maa.org/maa-invitational-competitions/ https://maa.org/student-programs/amc/ Abgerufen am 20. Oktober 2026.
McCarthy, John (2007): What Is Artificial Intelligence? Basic Questions. Stanford University, 12. November 2007. https://www-formal.stanford.edu/jmc/whatisai/node1.html https://web.archive.org/web/2025/https://www-formal.stanford.edu/jmc/whatisai/node1.html Abgerufen am 19. Oktober 2026.
McCarthy, John; Minsky, Marvin L.; Rochester, Nathaniel u. a. (1955): A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence. Abschrift auf der Website von John McCarthy, Stanford University, 31. August 1955. https://www-formal.stanford.edu/jmc/history/dartmouth/dartmouth.html https://web.archive.org/web/2025/http://www-formal.stanford.edu/jmc/history/dartmouth/dartmouth.html Abgerufen am 19. Oktober 2026.
McCulloch, Warren S.; Pitts, Walter (1943): A Logical Calculus of the Ideas Immanent in Nervous Activity. Bulletin of Mathematical Biophysics 5 (4), S. 115–133. https://doi.org/10.1007/BF02478259 Abgerufen am 19. Oktober 2026.
McCullough, Caleb (2026): Sam Altman says growing a single almond equals 38,000 ChatGPT queries. PolitiFact (Faktencheck, Urteil „Mostly False“), 4. September 2026. https://www.politifact.com/factchecks/2026/sep/04/sam-altman/chatgpt-almonds-ai-california-water-useage/ (Pressebericht) Abgerufen am 19. Oktober 2026.
Melis, Gábor; Dyer, Chris; Blunsom, Phil (2018): On the State of the Art of Evaluation in Neural Language Models. ICLR 2018 (Poster). https://arxiv.org/abs/1707.05589 Abgerufen am 19. Oktober 2026.
Meta (2024): Llama-3.1-8B-Instruct (Modellseite). Hugging Face, Juli 2024. https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct https://huggingface.co/unsloth/Llama-3.1-8B-Instruct/raw/main/config.json (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Meta (2025): The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation. Blogbeitrag, 5. April 2025. https://ai.meta.com/blog/llama-4-multimodal-intelligence/ (Herstellerangabe)
Meta Platforms (2026): Muse Spark Open: Modellkarte und Muse Model License. Hugging Face (Zugang nach Registrierung und Freigabe), 15. Oktober 2026. https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B (Herstellerangabe) Abgerufen am 2. November 2026.
METR (2026): Task-Completion Time Horizons of Frontier AI Models; Datei benchmark_results_1_1.yaml (Time Horizon 1.1). METR, Website (Hinweise vom 8.5. und 8.9.2026, seitdem nicht mehr gepflegt). Stand 25. Oktober 2026. https://metr.org/time-horizons/ https://metr.org/assets/benchmark_results_1_1.yaml Abgerufen am 25. Oktober 2026.
Metz, Rachel; Love, Julia (2024): Character.AI Co-Founders Hired by Google in Licensing Deal. Bloomberg, 2. August 2024. https://www.bloomberg.com/news/articles/2024-08-02/character-ai-co-founders-hired-by-google-in-licensing-deal (Pressebericht) Abgerufen am 26. Oktober 2026.
Meyes, Richard; Lu, Melanie; Waubert de Puiseau, Constantin u. a. (2019): Ablation Studies in Artificial Neural Networks. Preprint arXiv:1901.08644, 24. Januar 2019. https://arxiv.org/abs/1901.08644 Abgerufen am 25. Oktober 2026.
Microsoft (2026): 2026 Environmental Data Fact Sheet (Geschäftsjahr bis 30.6.2025). Microsoft, PDF (Tabellen 1A, 5, 8). https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/microsoft/msc/documents/presentations/CSR/2026-Microsoft-Environmental-Data-Fact-Sheet-PDF.pdf (Herstellerangabe) Abgerufen am 19. Oktober 2026.
★ Mikolov, Tomas; Chen, Kai; Corrado, Greg u. a. (2013): Efficient Estimation of Word Representations in Vector Space. ICLR 2013, Workshop Track. https://arxiv.org/abs/1301.3781 Abgerufen am 19. Oktober 2026.
Miller, Evan (2024): Adding Error Bars to Evals: A Statistical Approach to Language Model Evaluations. Preprint, 1. November 2024. https://arxiv.org/abs/2411.00640 Abgerufen am 19. Oktober 2026.
Mistral AI (2023): Mixtral of experts. Mistral AI, News (Veröffentlichung der Gewichte von Mixtral 8x7B), 11. Dezember 2023. https://mistral.ai/news/mixtral-of-experts (Herstellerangabe) Abgerufen am 20. Oktober 2026.
Mistral AI (2025): Our contribution to a global environmental standard for AI. Mistral AI, Blog, 22. Juli 2025. https://mistral.ai/news/our-contribution-to-a-global-environmental-standard-for-ai/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Mistral AI (2026a): Le Chat is now Vibe; Training-, Opt-out-, Aufbewahrungs- und Memories-Seiten; Mistral Docs „Vibe“. Mistral Help Center; Mistral Docs. https://help.mistral.ai/en/articles/682992-le-chat-is-now-vibe https://help.mistral.ai/en/articles/347617-do-you-use-my-user-data-to-train-your-artificial-intelligence-models https://help.mistral.ai/en/articles/455207-can-i-opt-out-of-my-input-or-output-data-being-used-for-training https://help.mistral.ai/en/articles/347612-can-i-activate-zero-data-retention-zdr https://help.mistral.ai/en/articles/347632-can-people-other-than-myself-view-my-conversations https://help.mistral.ai/en/articles/396497-how-do-you-handle-my-data-when-using-the-memories-feature https://docs.mistral.ai/vibe (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Mistral AI (2026b): Pricing. mistral.ai (API-Preise, FAQ, Tarife Free, Pro, Team, Enterprise). https://mistral.ai/pricing (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Mistral AI (2026c): Introducing Mistral Large 4. Mistral AI, Blog, 6. Oktober 2026. https://mistral.ai/news/mistral-large-4 (Herstellerangabe) Abgerufen am 2. November 2026.
Mistral AI (2026d): Mistral Large 4 (Modellseite). Mistral-Dokumentation (Public Preview), 6. Oktober 2026. https://docs.mistral.ai/models/mistral-large-4-0 (Herstellerangabe) Abgerufen am 2. November 2026.
Mistral AI (2026e): Mistral-Large-4-1T-A52B: Modellkarte und Mistral Ecosystem Open-Weight License. Hugging Face, 29. Oktober 2026. https://huggingface.co/mistralai/Mistral-Large-4-1T-A52B (Herstellerangabe) Abgerufen am 2. November 2026.
MIT Department of Economics (2025): Assuring an accurate research record. MIT Economics, News, 16. Mai 2025. https://economics.mit.edu/news/assuring-accurate-research-record Abgerufen am 25. Oktober 2026.
Mitchell, Margaret u. a. (2019): Model Cards for Model Reporting. FAT* 2019; Preprint arXiv:1810.03993 (v1 5.10.2018). https://arxiv.org/abs/1810.03993 Abgerufen am 19. Oktober 2026.
Modarressi, Ali; Deilamsalehy, Hanieh; Dernoncourt, Franck u. a. (2025): NoLiMa: Long-Context Evaluation Beyond Literal Matching. ICML 2025; Preprint arXiv:2502.05167 (v1 vom 7.2.2025, v3 vom 9.7.2025; LMU München, Adobe Research), 7. Februar 2025. https://arxiv.org/abs/2502.05167 Abgerufen am 23. Oktober 2026.
Model Context Protocol (2026): Specification, Version 2026-07-28. modelcontextprotocol.io, 28. Juli 2026. https://modelcontextprotocol.io/specification/latest Abgerufen am 19. Oktober 2026.
Muennighoff, Niklas u. a. (2024): OLMoE: Open Mixture-of-Experts Language Models. Preprint, arXiv:2409.02060 (zitiert nach der Fassung vom 3.3.2025), 3. September 2024. https://arxiv.org/abs/2409.02060 Abgerufen am 20. Oktober 2026.
Murphy, Brendan u. a. (2025): Illusory Safety: Redteaming DeepSeek R1 and the Strongest Fine-Tunable Models of OpenAI, Anthropic, and Google. FAR.AI, Blog, 3. Februar 2025. https://www.far.ai/blog/illusory-safety-redteaming-deepseek-r1-and-the-strongest-fine-tunable-models-of-openai-anthropic-and-google Abgerufen am 19. Oktober 2026.
Nanda, Neel (2023): Actually, Othello-GPT Has A Linear Emergent World Representation. Blog neelnanda.io, 28. März 2023. https://www.neelnanda.io/mechanistic-interpretability/othello Abgerufen am 25. Oktober 2026.
Nanda, Neel; Lee, Andrew; Wattenberg, Martin (2023): Emergent Linear Representations in World Models of Self-Supervised Sequence Models. BlackboxNLP@EMNLP 2023; Preprint arXiv:2309.00941 (v2 vom 7.9.2023). https://arxiv.org/abs/2309.00941 Abgerufen am 25. Oktober 2026.
National Academies of Sciences, Engineering, and Medicine (2019): Reproducibility and Replicability in Science (Summary). The National Academies Press, Washington, DC. https://www.nationalacademies.org/read/25303/chapter/2 Abgerufen am 25. Oktober 2026.
National Institute of Standards and Technology (NIST) (2026): NIST Digital Library of Mathematical Functions, Kap. 4 „Elementary Functions“, §4.2 „Definitions“ und §4.8 „Identities“. Online, Version 1.2.8 vom 15.9.2026, 15. September 2026. https://dlmf.nist.gov/4.2 https://dlmf.nist.gov/4.8 Abgerufen am 24. Oktober 2026.
National Technical Committee 260 on Cybersecurity (TC260) (2024): Basic Safety Requirements for Generative Artificial Intelligence Services (TC260-003). TC260, 29. Februar 2024; englische Übersetzung des Center for Security and Emerging Technology (CSET), 4. April 2024. https://cset.georgetown.edu/publication/china-safety-requirements-for-generative-ai-final/ https://cset.georgetown.edu/wp-content/uploads/t0588_generative_AI_safety_EN.pdf Abgerufen am 19. Oktober 2026.
nativ-community (2026): Kolibri-1-MLX-4bit. Hugging Face, Repository angelegt am 3.10.2026, 3. Oktober 2026. https://huggingface.co/nativ-community/Kolibri-1-MLX-4bit (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Nature (2025): Bring us your LLMs: why peer review is good for AI models. Nature 645, S. 559 (Editorial), 17. September 2025. https://www.nature.com/articles/d41586-025-02979-9 Abgerufen am 25. Oktober 2026.
NeurIPS (2020): Announcing the NeurIPS 2020 award recipients. NeurIPS Blog, 7. Dezember 2020. https://blog.neurips.cc/2020/12/07/announcing-the-neurips-2020-award-recipients/ Abgerufen am 19. Oktober 2026.
NeurIPS (2022): Announcing the NeurIPS 2022 Awards. NeurIPS Blog, 21. November 2022. https://blog.neurips.cc/2022/11/21/announcing-the-neurips-2022-awards/ Abgerufen am 19. Oktober 2026.
NeurIPS (2023): Announcing the NeurIPS 2023 Paper Awards. NeurIPS Blog, 11. Dezember 2023. https://blog.neurips.cc/2023/12/11/announcing-the-neurips-2023-paper-awards/ Abgerufen am 25. Oktober 2026.
NeurIPS (2026a): NeurIPS 2026 Main Track Handbook. neurips.cc. https://neurips.cc/Conferences/2026/MainTrackHandbook Abgerufen am 19. Oktober 2026.
NeurIPS (2026b): NeurIPS Paper Checklist Guidelines. neurips.cc. https://neurips.cc/public/guides/PaperChecklist Abgerufen am 19. Oktober 2026.
NeurIPS 2025 Program Committee Chairs (2025): Reflections on the 2025 Review Process from the Program Committee Chairs. NeurIPS Blog, 30. September 2025. https://blog.neurips.cc/2025/09/30/reflections-on-the-2025-review-process-from-the-program-committee-chairs/ Abgerufen am 19. Oktober 2026.
Newton, Casey (2025): Grok’s new porn companion is rated for kids 12+ in the App Store. Platformer (Newsletter), 15. Juli 2025. https://www.platformer.news/grok-ani-app-store-rating-nsfw-avatar-apple (Pressebericht) Abgerufen am 25. Oktober 2026.
Neyman, Jerzy (1934): On the Two Different Aspects of the Representative Method: The Method of Stratified Sampling and the Method of Purposive Selection. Journal of the Royal Statistical Society 97 (4), S. 558–625. https://doi.org/10.2307/2342192 Abgerufen am 25. Oktober 2026.
Nicoletti, Leonardo; Bass, Dina (2023): Humans Are Biased. Generative AI Is Even Worse. Bloomberg, Technology + Equality, 8. Juni 2023. https://www.bloomberg.com/graphics/2023-generative-ai-bias/ (Pressebericht) Abgerufen am 25. Oktober 2026.
NIST (o. J.): Dataplot Reference Manual: SIGN TEST; PROPORTION CONFIDENCE LIMITS. National Institute of Standards and Technology. https://www.itl.nist.gov/div898/software/dataplot/refman1/auxillar/signtest.htm https://itl.nist.gov/div898/software/dataplot/refman1/auxillar/propconf.htm Abgerufen am 19. Oktober 2026.
NIST Center for AI Standards and Innovation (CAISI) (2025): Evaluation of DeepSeek AI Models. NIST, Gaithersburg, 30. September 2025, aktualisiert am 20. November 2025. https://www.nist.gov/system/files/documents/2025/09/30/CAISI_Evaluation_of_DeepSeek_AI_Models.pdf https://www.nist.gov/news-events/news/2025/09/caisi-evaluation-deepseek-ai-models-finds-shortcomings-and-risks Abgerufen am 19. Oktober 2026.
NIST/SEMATECH (o. J. a): e-Handbook of Statistical Methods, Abschn. 7.1.4: What are confidence intervals? National Institute of Standards and Technology. https://www.itl.nist.gov/div898/handbook/prc/section1/prc14.htm Abgerufen am 25. Oktober 2026.
NIST/SEMATECH (o. J. b): e-Handbook of Statistical Methods: Confidence intervals. National Institute of Standards and Technology, Abschn. 7.2.4.1. https://www.itl.nist.gov/div898/handbook/prc/section2/prc241.htm Abgerufen am 19. Oktober 2026.
noyb – European Center for Digital Rights (2025): AI hallucinations: ChatGPT created a fake child murderer. Wien (Beschwerdeführerin); dazu Fallseite „C097 OpenAI OpCo, LLC“, 20. März 2025. https://noyb.eu/en/ai-hallucinations-chatgpt-created-fake-child-murderer https://noyb.eu/en/project/artificial-intelligence/c097 Abgerufen am 19. Oktober 2026.
Nummer gegen Kummer e. V. (2026): Kinder- und Jugendtelefon 116 111. Website. https://www.nummergegenkummer.de/ Abgerufen am 2. November 2026.
NVIDIA (2025a): GeForce RTX 5090 und GeForce RTX 5060 Family (Produktseiten). nvidia.com. https://www.nvidia.com/en-us/geforce/graphics-cards/50-series/rtx-5090/ https://www.nvidia.com/en-us/geforce/graphics-cards/50-series/rtx-5060-family/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
NVIDIA (2025b): NVIDIA RTX Blackwell GPU Architecture. Whitepaper, Version 1.1. https://images.nvidia.com/aem-dam/Solutions/geforce/blackwell/nvidia-rtx-blackwell-gpu-architecture.pdf (Herstellerangabe) Abgerufen am 19. Oktober 2026.
NVIDIA (2026a): Form 10-K für das Geschäftsjahr bis 25. Januar 2026. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1045810/000104581026000021/nvda-20260125.htm Abgerufen am 19. Oktober 2026.
NVIDIA (2026b): NVIDIA Announces Financial Results for Second Quarter Fiscal 2027. Pressemitteilung, Exhibit 99.1, U.S. Securities and Exchange Commission, 26. August 2026. https://www.sec.gov/Archives/edgar/data/0001045810/000104581026000073/q2fy27pr.htm Abgerufen am 19. Oktober 2026.
NVIDIA (2026c): CUDA Programming Guide, Version 13.4.2. NVIDIA Developer Documentation, Kap. 1.1 „Introduction“, 10. September 2026. https://docs.nvidia.com/cuda/cuda-programming-guide/01-introduction/introduction.html (Herstellerangabe) Abgerufen am 19. Oktober 2026.
NVIDIA (o. J. a): GB200 NVL72. Produktseite. https://www.nvidia.com/en-us/data-center/gb200-nvl72/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
NVIDIA (o. J. b): GeForce Graphics Cards – Compare. nvidia.com. https://www.nvidia.com/en-us/geforce/graphics-cards/compare/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
NVIDIA (o. J. c): GPU Performance Background User’s Guide. NVIDIA Deep Learning Performance Documentation. https://docs.nvidia.com/deeplearning/performance/dl-performance-gpu-background/index.html (Herstellerangabe) Abgerufen am 20. Oktober 2026.
NVIDIA (o. J. d): NVIDIA DGX Spark. Produktseite mit Spezifikationen (128 GB, 273 GB/s). https://www.nvidia.com/en-us/products/workstations/dgx-spark/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
NVIDIA (o. J. e): NVIDIA H100 Tensor Core GPU. Produktseite (H100 SXM: 80 GB, 3,35 TB/s). https://www.nvidia.com/en-us/data-center/h100/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
NVIDIA (o. J. f): NVIDIA H200 Tensor Core GPU. Produktseite (Angaben als vorläufig gekennzeichnet). https://www.nvidia.com/en-us/data-center/h200/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
O’Brien, Kyle; Casper, Stephen; Anthony, Quentin u. a. (2025): Deep Ignorance: Filtering Pretraining Data Builds Tamper-Resistant Safeguards into Open-Weight LLMs. Preprint, arXiv:2508.06601, 8. August 2025; Fassung 2 vom 17. Februar 2026. https://arxiv.org/abs/2508.06601 Abgerufen am 20. Oktober 2026.
Oelen, Allard; Jaradeh, Mohamad Yaser; Auer, Sören (2024): ORKG ASK: a Neuro-symbolic Scholarly Search and Exploration System. Preprint, arXiv:2412.04977 (TIB Hannover, L3S), 6. Dezember 2024. https://arxiv.org/abs/2412.04977 Abgerufen am 20. Oktober 2026.
Öko-Institut (o. J.): Künstliche Intelligenz und Software (Themenseite). Öko-Institut, Freiburg. https://www.oeko.de/themen/digitalisierung/kuenstliche-intelligenz-und-software/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
olegshulyakov u. a. (2025): Performance of llama.cpp on Nvidia CUDA. GitHub-Diskussion Nr. 15013, ggml-org/llama.cpp, seit 1.8.2025 laufend ergänzt (Community-Messungen), 1. August 2025. https://github.com/ggml-org/llama.cpp/discussions/15013 (Sekundärquelle) Abgerufen am 20. Oktober 2026.
Olshausen, Bruno A.; Field, David J. (1996): Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature 381, S. 607–609, 13. Juni 1996. https://doi.org/10.1038/381607a0 Abgerufen am 25. Oktober 2026.
Open Source Initiative (2024): The Open Source AI Definition – 1.0. opensource.org (mit Ankündigung vom selben Tag), 28. Oktober 2024. https://opensource.org/ai/open-source-ai-definition https://opensource.org/blog/the-open-source-initiative-announces-the-release-of-the-industrys-first-open-source-ai-definition Abgerufen am 19. Oktober 2026.
Open Source Initiative (o. J.): Answers to Frequently Asked Questions (Open Source AI Definition). opensource.org. https://opensource.org/ai/faq Abgerufen am 19. Oktober 2026.
OpenAI (2022): Introducing ChatGPT. OpenAI, Blog, 30. November 2022. https://openai.com/index/chatgpt/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2023): GPT-4 Technical Report. Preprint, arXiv:2303.08774 (v6 vom 4.3.2024), 15. März 2023. https://arxiv.org/abs/2303.08774 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2024a): Video generation models as world simulators. OpenAI, Research, 15. Februar 2024. https://openai.com/index/video-generation-models-as-world-simulators/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2024b): Understanding the source of what we see and hear online. OpenAI, Blog (Nachtrag vom August 2024 zu Text-Wasserzeichen), 7. Mai 2024. https://openai.com/index/understanding-the-source-of-what-we-see-and-hear-online/ (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2024c): Introducing the Model Spec. OpenAI, Blog, 8. Mai 2024. https://openai.com/index/introducing-the-model-spec/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
OpenAI (2024d): How the voices for ChatGPT were chosen. OpenAI, Blog (Ergänzungen vom 20. und 22.5.2024), 19. Mai 2024. https://openai.com/index/how-the-voices-for-chatgpt-were-chosen (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2024e): GPT-4o System Card. Preprint arXiv:2410.21276 (25.10.2024), 8. August 2024. https://arxiv.org/abs/2410.21276 (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2024f): Introducing SWE-bench Verified. Blogbeitrag, 13. August 2024. https://openai.com/index/introducing-swe-bench-verified/ (Herstellerangabe)
OpenAI (2024g): Learning to reason with LLMs. OpenAI, Blog, 12. September 2024. https://openai.com/index/learning-to-reason-with-llms/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2024h): Introducing ChatGPT Pro. openai.com (Archivfassung vom 5.12.2024), 5. Dezember 2024. https://openai.com/index/introducing-chatgpt-pro/ https://web.archive.org/web/20241205191841/https://openai.com/index/introducing-chatgpt-pro/ (Herstellerangabe) Abgerufen am 26. Oktober 2026.
OpenAI (2025a): Disrupting malicious uses of AI: June 2025. OpenAI, Threat Intelligence Report (PDF). https://cdn.openai.com/threat-intelligence-reports/5f73af09-a3a3-4a55-992e-069237681620/disrupting-malicious-uses-of-ai-june-2025.pdf (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2025b): Introducing data residency in Europe. San Francisco (Updates bis 16.1.2026), 5. Februar 2025. https://openai.com/index/introducing-data-residency-in-europe/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
OpenAI (2025c): Sycophancy in GPT-4o: What happened and what we’re doing about it. OpenAI, Blog, 29. April 2025. https://openai.com/index/sycophancy-in-gpt-4o/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2025d): Expanding on what we missed with sycophancy. OpenAI, Blog, 2. Mai 2025. https://openai.com/index/expanding-on-sycophancy/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2025e): gpt-oss-120b & gpt-oss-20b Model Card. Preprint, arXiv:2508.10925; dazu die Modellseite auf Hugging Face (Repository vom 4.8.2025), 8. August 2025. https://arxiv.org/abs/2508.10925 https://huggingface.co/openai/gpt-oss-120b (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2025f): GPT-5 System Card. PDF, 13. August 2025. https://cdn.openai.com/gpt-5-system-card.pdf (Herstellerangabe)
OpenAI (2025g): Collective alignment: public input on our Model Spec. OpenAI, Blog, 27. August 2025. https://openai.com/index/collective-alignment-aug-2025-updates (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2025h): Why language models hallucinate. Blogbeitrag, 5. September 2025. https://openai.com/index/why-language-models-hallucinate/ (Herstellerangabe)
OpenAI (2025i): Strengthening ChatGPT’s responses in sensitive conversations. OpenAI, Blog, 27. Oktober 2025. https://openai.com/index/strengthening-chatgpt-responses-in-sensitive-conversations/ (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2025j): Continuously hardening ChatGPT Atlas against prompt injection attacks. OpenAI Blog, 22. Dezember 2025. https://openai.com/index/hardening-atlas-against-prompt-injection/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
OpenAI (2026a): Deprecations. OpenAI-API-Dokumentation. https://developers.openai.com/api/docs/deprecations (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026b): GPT-6 Astra (Modellseite) und Models (Übersicht). OpenAI-API-Dokumentation. https://developers.openai.com/api/docs/models/gpt-6-astra https://developers.openai.com/api/docs/models (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026c): How your data is used to improve model performance; Chat and file retention in ChatGPT; How OpenAI handles data in consumer services; Memory FAQ. OpenAI Help Center. https://help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance https://help.openai.com/en/articles/8983778-chat-and-file-retention-policies-in-chatgpt https://help.openai.com/en/articles/7039943-how-openai-handles-data-in-consumer-services https://help.openai.com/en/articles/8590148-memory-faq (Herstellerangabe) Abgerufen am 19. Oktober 2026.
OpenAI (2026d): Pricing. OpenAI API-Dokumentation (Standard-Tarif, Textmodelle). https://developers.openai.com/api/docs/pricing (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026e): Pricing. ChatGPT Learn (früher OpenAI-Entwicklerdokumentation, Codex), Pläne Free, Go, Plus, Pro, Business. https://learn.chatgpt.com/docs/pricing https://developers.openai.com/codex/pricing (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026f): Reasoning models. OpenAI-API-Dokumentation (ohne Datum). https://developers.openai.com/api/docs/guides/reasoning (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2026g): Temporary chat in ChatGPT. OpenAI Help Center (Stand 31.10.2026). https://help.openai.com/en/articles/8914046-temporary-chat-in-chatgpt (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026h): Using GPT-6. OpenAI-API-Dokumentation, Leitfaden zum Umstieg (/guides/latest-model); dazu API-Referenz „Create a model response“. https://developers.openai.com/api/docs/guides/latest-model https://developers.openai.com/api/reference/resources/responses/methods/create (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026i): What is ChatGPT Plus? OpenAI Help Center (Archivfassung vom 14.10.2026). https://help.openai.com/en/articles/6950777-what-is-chatgpt-plus https://web.archive.org/web/20261014181134/https://help.openai.com/en/articles/6950777-what-is-chatgpt-plus (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2026j): What to know about the Sora discontinuation. OpenAI Help Center. https://help.openai.com/en/articles/20001152-what-to-know-about-the-sora-discontinuation (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026k): Enterprise privacy at OpenAI. OpenAI, Website (aktualisiert 8.1.2026), 8. Januar 2026. https://openai.com/enterprise-privacy/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
OpenAI (2026l): Retiring GPT‑4o, GPT‑4.1, GPT‑4.1 mini, and OpenAI o4-mini in ChatGPT. OpenAI, Blog, 29. Januar 2026. https://openai.com/index/retiring-gpt-4o-and-older-models (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2026m): Updated Stakes for American-Led, Democratic AI. Memo an das U.S. House Select Committee on Strategic Competition between the United States and the Chinese Communist Party; veröffentlicht über Bloomberg, 12. Februar 2026. https://assets.bwbx.io/documents/users/iqjWHBFdfxIU/rRmql_jJcxb4/v0 (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2026n): Why SWE-bench Verified no longer measures frontier coding capabilities. Blogbeitrag, 23. Februar 2026. https://openai.com/index/why-we-no-longer-evaluate-swe-bench-verified/ (Herstellerangabe)
OpenAI (2026o): Separating signal from noise in coding evaluations. Blogbeitrag, 8. Juli 2026. https://openai.com/index/separating-signal-from-noise-coding-evaluations/ (Herstellerangabe)
OpenAI (2026p): The Hugging Face incident and the road ahead. OpenAI, Blog (mit technischem Bericht), 26. August 2026. https://openai.com/index/hugging-face-incident-and-the-road-ahead (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2026q): GPT-6 Astra System Card. OpenAI Deployment Safety Hub (Änderungen 9.9., 22.9. und 29.9.2026), 3. September 2026. https://deploymentsafety.openai.com/gpt-6-astra (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (2026r): Research acceleration: The view inside OpenAI. OpenAI, Blog, 6. September 2026. https://openai.com/index/research-acceleration-view-inside-openai (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2026s): Addendum to GPT-6 Astra System Card: GPT-6.1 Sol. OpenAI Deployment Safety Hub, 29. September 2026. https://deploymentsafety.openai.com/gpt-6-1-sol (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2026t): Our approach to EU text provenance rules. OpenAI, Blog; dazu technischer Bericht „textGrain: Entropy-Calibrated Watermarking for Language Model Text“, 5. Oktober 2026. https://openai.com/index/eu-text-provenance (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (2026u): GPT‑6 and Intelligent UI for everyone. OpenAI, Blog, 7. Oktober 2026. https://openai.com/index/gpt-6-for-everyone/ (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI (2026v): Disrupting AI-enabled „false front“ operations. OpenAI, Blog (Threat Intelligence), 8. Oktober 2026. https://openai.com/index/disrupting-ai-enabled-false-front-operations/ (Herstellerangabe) Abgerufen am 25. Oktober 2026.
OpenAI (o. J. a): Europe Terms of Use; Terms of Use (Fassung für die übrigen Länder). OpenAI, Richtlinien. https://openai.com/policies/eu-terms-of-use/ https://openai.com/policies/row-terms-of-use/ (Herstellerangabe) Abgerufen am 20. Oktober 2026.
OpenAI (o. J. b): OpenAI Evals (Repository) und Evaluation best practices (API-Dokumentation). GitHub; developers.openai.com. https://github.com/openai/evals https://developers.openai.com/api/docs/guides/evaluation-best-practices (Herstellerangabe) Abgerufen am 19. Oktober 2026.
OpenAI (o. J. c): tiktoken, Datei tiktoken_ext/openai_public.py. Quellcode-Repositorium, GitHub. https://github.com/openai/tiktoken/blob/main/tiktoken_ext/openai_public.py Abgerufen am 19. Oktober 2026.
OpenAI (o. J. d): Tokenizer. Web-Werkzeug der OpenAI-Plattform. https://platform.openai.com/tokenizer (Herstellerangabe) Abgerufen am 2. November 2026.
OpenAI; Hugging Face (o. J.): openai-community/gpt2, Datei config.json. Modellrepositorium, Hugging Face. https://huggingface.co/openai-community/gpt2/blob/main/config.json Abgerufen am 19. Oktober 2026.
OpenGPT-X (2024): Teuken-7B-instruct-research-v0.4, Model Card und config.json. Hugging Face. https://huggingface.co/openGPT-X/Teuken-7B-instruct-research-v0.4 (Herstellerangabe) Abgerufen am 19. Oktober 2026.
OrcaRouter (2026): Kolibri (chinesische Fassung). Blogbeitrag. Stand 13. Oktober 2026. https://www.orcarouter.ai/zh-CN/blog/kolibri (Sekundärquelle)
★ Ouyang, Long u. a. (2022): Training language models to follow instructions with human feedback. NeurIPS 2022. https://arxiv.org/abs/2203.02155 Abgerufen am 19. Oktober 2026.
OWASP Gen AI Security Project (2025): LLM01:2025 Prompt Injection (OWASP Top 10 for LLM Applications 2025). genai.owasp.org. https://genai.owasp.org/llmrisk/llm01-prompt-injection/ (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Panickssery, Arjun; Bowman, Samuel R.; Feng, Shi (2024): LLM Evaluators Recognize and Favor Their Own Generations. NeurIPS 2024. https://arxiv.org/abs/2404.13076 Abgerufen am 19. Oktober 2026.
Patterson, David u. a. (2021): Carbon Emissions and Large Neural Network Training. Preprint, arXiv:2104.10350v3 (Google und University of California, Berkeley), 23. April 2021. https://arxiv.org/abs/2104.10350 Abgerufen am 19. Oktober 2026.
Pfister, Jan; Wunderle, Julia; Hotho, Andreas (2025): LLäMmlein: Transparent, Compact and Competitive German-Only Language Models from Scratch. Proceedings of the 63rd Annual Meeting of the ACL (Volume 1: Long Papers), Wien, S. 2227–2246, DOI 10.18653/v1/2025.acl-long.111; Preprint arXiv:2411.11171 (v1 17.11.2024), Juli 2025. https://aclanthology.org/2025.acl-long.111/ https://arxiv.org/abs/2411.11171 Abgerufen am 19. Oktober 2026.
Pfungst, Oskar (1907): Das Pferd des Herrn von Osten (Der kluge Hans). Ein Beitrag zur experimentellen Tier- und Menschen-Psychologie. Mit einer Einleitung von C. Stumpf. Leipzig: Johann Ambrosius Barth. https://archive.org/details/daspferddesherr00stumgoog Abgerufen am 19. Oktober 2026.
Pfungst, Oskar (1911): Clever Hans (The Horse of Mr. Von Osten). A Contribution to Experimental Animal and Human Psychology. Übersetzt von Carl L. Rahn. New York: Henry Holt and Company (englische Übersetzung von Pfungst 1907). https://www.gutenberg.org/ebooks/33936 Abgerufen am 19. Oktober 2026.
Phan, Long u. a. (2026): A benchmark of expert-level academic questions to assess AI capabilities (Humanity’s Last Exam). Nature 649, 1139–1146; Preprint arXiv:2501.14249 (Fundstellen in Kap. 8 und 9 nach v12 vom 19.10.2026). https://arxiv.org/abs/2501.14249 https://www.nature.com/articles/s41586-025-09962-4
Phang, Jason u. a. (2025): Investigating Affective Use and Emotional Well-being on ChatGPT. OpenAI und MIT Media Lab, Bericht vom 21.3.2025; Preprint arXiv:2504.03888 (4.4.2025), 21. März 2025. https://arxiv.org/abs/2504.03888 (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Physical Intelligence (2025a): π0.6: a VLA that Learns from Experience*. Physical Intelligence, Blog, 17. November 2025. https://www.pi.website/blog/pistar06 (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Physical Intelligence (2025b): Moravec’s Paradox and the Robot Olympics. Physical Intelligence, Blog, 22. Dezember 2025. https://www.pi.website/blog/olympics (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Pineau, Joelle u. a. (2021): Improving Reproducibility in Machine Learning Research (A Report from the NeurIPS 2019 Reproducibility Program). Journal of Machine Learning Research 22 (164): 1–20. https://jmlr.org/papers/v22/20-303.html https://arxiv.org/abs/2003.12206 Abgerufen am 19. Oktober 2026.
Poley, Christoph; Uhlmann, Sandro; Busse, Frank u. a. (2025): Automatic Subject Cataloguing at the German National Library. LIBER Quarterly 35 (1), S. 1–29, DOI 10.53377/lq.19422, 8. April 2025. https://liberquarterly.eu/article/view/19422 Abgerufen am 19. Oktober 2026.
princeton-nlp (o. J.): SWE-bench Verified (Datensatz). Hugging Face. https://huggingface.co/datasets/princeton-nlp/SWE-bench_Verified
PyTorch Foundation (o. J. a): CUDA semantics. PyTorch-Dokumentation, Version 2.14. https://docs.pytorch.org/docs/2.14/notes/cuda.html Abgerufen am 19. Oktober 2026.
PyTorch Foundation (o. J. b): torch.nn.KLDivLoss. PyTorch-Dokumentation, Version 2.14. https://docs.pytorch.org/docs/2.14/generated/torch.nn.KLDivLoss.html Abgerufen am 25. Oktober 2026.
PyTorch Foundation (o. J. c): torch.nn.Sigmoid. PyTorch-Dokumentation, Version 2.14. https://docs.pytorch.org/docs/2.14/generated/torch.nn.Sigmoid.html Abgerufen am 25. Oktober 2026.
Qi, Xiangyu; Panda, Ashwinee; Lyu, Kaifeng u. a. (2025): Safety Alignment Should Be Made More Than Just a Few Tokens Deep. ICLR 2025; Preprint arXiv:2406.05946 (v1 vom 10.6.2024). https://arxiv.org/abs/2406.05946 Abgerufen am 19. Oktober 2026.
Qi, Xiangyu u. a. (2024): On Evaluating the Durability of Safeguards for Open-Weight LLMs. Preprint arXiv:2412.07097, 10. Dezember 2024. https://arxiv.org/abs/2412.07097 Abgerufen am 25. Oktober 2026.
Qi, Xiangyu; Zeng, Yi; Xie, Tinghao u. a. (2023): Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To! ICLR 2024; Preprint arXiv:2310.03693, 5. Oktober 2023. https://arxiv.org/abs/2310.03693 Abgerufen am 2. November 2026.
Qwen-Team (Alibaba) (2025): Qwen3Guard Technical Report; Modellkarte Qwen3Guard-Gen-8B. Technischer Bericht, Preprint arXiv:2510.14276, 16. Oktober 2025; Modelle auf Hugging Face seit 23. September 2025. https://arxiv.org/abs/2510.14276 https://huggingface.co/Qwen/Qwen3Guard-Gen-8B (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Radford, Alec; Kim, Jong Wook; Xu, Tao u. a. (2022): Robust Speech Recognition via Large-Scale Weak Supervision. Preprint arXiv:2212.04356 (OpenAI), 6. Dezember 2022. https://arxiv.org/abs/2212.04356 Abgerufen am 25. Oktober 2026.
Radford, Alec; Narasimhan, Karthik; Salimans, Tim u. a. (2018): Improving Language Understanding by Generative Pre-Training. Technischer Bericht, OpenAI (nicht begutachtet). https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf Abgerufen am 19. Oktober 2026.
Radford, Alec u. a. (2021): Learning Transferable Visual Models From Natural Language Supervision. Preprint arXiv:2103.00020 (OpenAI), 26. Februar 2021. https://arxiv.org/abs/2103.00020 Abgerufen am 20. Oktober 2026.
Radford, Alec; Wu, Jeffrey; Child, Rewon u. a. (2019): Language Models are Unsupervised Multitask Learners. Technischer Bericht, OpenAI (Ankündigung 14.2.2019), 14. Februar 2019. https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf Abgerufen am 19. Oktober 2026.
★ Rafailov, Rafael u. a. (2023): Direct Preference Optimization: Your Language Model is Secretly a Reward Model. NeurIPS 2023. https://arxiv.org/abs/2305.18290 Abgerufen am 19. Oktober 2026.
Raff, Edward (2019): A Step Toward Quantifying Independently Reproducible Machine Learning Research. NeurIPS 2019; Preprint arXiv:1909.06674, 14. September 2019. https://arxiv.org/abs/1909.06674 Abgerufen am 25. Oktober 2026.
Rajbhandari, Samyam; Rasley, Jeff; Ruwase, Olatunji u. a. (2020): ZeRO: Memory Optimizations Toward Training Trillion Parameter Models. SC20; Preprint arXiv:1910.02054v3, 13. Mai 2020. https://arxiv.org/abs/1910.02054 Abgerufen am 19. Oktober 2026.
Region Skåne (2021): Mammography Screening With Artificial Intelligence (MASAI). ClinicalTrials.gov, NCT04838756 (Studienregister; Studienleiterin Kristina Lång), 9. April 2021. https://clinicaltrials.gov/study/NCT04838756 Abgerufen am 29. Oktober 2026.
Rein, David u. a. (2023): GPQA: A Graduate-Level Google-Proof Q&A Benchmark. Preprint. https://arxiv.org/abs/2311.12022
Reuters (2025): Nvidia stock plunges as DeepSeek sparks AI stock selloff. Reuters, Meldung vom 27./28.1.2025, gelesen in der Übernahme durch Khaleej Times, 28. Januar 2025. https://www.khaleejtimes.com/business/tech/nvidia-stock-plunges-as-deepseek-sparks-ai-stock-selloff-sends-shockwaves (Pressebericht) Abgerufen am 25. Oktober 2026.
RIKEN AIP (2026b): 81 Papers Accepted for NeurIPS 2026. RIKEN Center for Advanced Intelligence Project, News, 14. Oktober 2026. https://aip.riken.jp/news/neurips2026/ (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Rombach, Robin; Blattmann, Andreas; Lorenz, Dominik u. a. (2021): High-Resolution Image Synthesis with Latent Diffusion Models. Preprint arXiv:2112.10752 (v2 vom 13.4.2022), 20. Dezember 2021. https://arxiv.org/abs/2112.10752 Abgerufen am 20. Oktober 2026.
Rosalsky, Greg (2025): Why the AI world is suddenly obsessed with a 160-year-old economics paradox. NPR, Planet Money, 4. Februar 2025. https://www.npr.org/sections/planet-money/2025/02/04/g-s1-46018/ai-deepseek-economics-jevons-paradox https://www.nprillinois.org/2025-02-04/why-the-ai-world-is-suddenly-obsessed-with-a-160-year-old-economics-paradox (Pressebericht) Abgerufen am 19. Oktober 2026.
★ Rumelhart, David E.; Hinton, Geoffrey E.; Williams, Ronald J. (1986): Learning representations by back-propagating errors. Nature 323, 533–536, 9. Oktober 1986. https://doi.org/10.1038/323533a0 Abgerufen am 19. Oktober 2026.
Rumi (2025): New ChatGPT Models Seem to Leave Watermarks on Text. Rumi, Newsroom (Nachträge vom 22. und 23.4.2025 mit der Antwort von OpenAI), 20. April 2025. https://www.rumidocs.com/newsroom/new-chatgpt-models-seem-to-leave-watermarks-on-text (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Scale AI (2026): Humanity’s Last Exam Leaderboard. Website. Stand 17. September 2026 (abgerufen 2. November 2026). https://labs.scale.com/leaderboard/humanitys_last_exam Abgerufen am 2. November 2026.
Schick, Timo; Dwivedi-Yu, Jane; Dessì, Roberto u. a. (2023): Toolformer: Language Models Can Teach Themselves to Use Tools. Preprint arXiv:2302.04761 (v1 9.2.2023); NeurIPS 2023 (dort zusätzlich Eric Hambro), 9. Februar 2023. https://arxiv.org/abs/2302.04761 https://proceedings.neurips.cc/paper_files/paper/2023/hash/d842425e4bf79ba039352da0f658a906-Abstract-Conference.html Abgerufen am 19. Oktober 2026.
Schoon, Ben (2026): Gemini’s co-lead is leaving Google to join OpenAI. 9to5Google, 17. Juni 2026. https://9to5google.com/2026/06/17/geminis-co-lead-is-leaving-google-to-join-openai/ (Pressebericht) Abgerufen am 2. November 2026.
Schulman, John; Thinking Machines Lab (2025): LoRA Without Regret. Thinking Machines Lab: Connectionism (Blog), DOI 10.64434/tml.20250929, 29. September 2025. https://thinkingmachines.ai/blog/lora/ (Herstellerangabe) Abgerufen am 24. Oktober 2026.
Schulman, John; Wolski, Filip; Dhariwal, Prafulla u. a. (2017): Proximal Policy Optimization Algorithms. Preprint arXiv:1707.06347 (OpenAI; v1 vom 20.7.2017, zitiert nach v2 vom 28.8.2017), 20. Juli 2017. https://arxiv.org/abs/1707.06347 Abgerufen am 24. Oktober 2026.
Science Media Center Germany (o. J.): Science Media Center Germany. Website, Köln. https://www.sciencemediacenter.de/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
Sclar, Melanie u. a. (2024): Quantifying Language Models’ Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formatting. ICLR 2024. https://arxiv.org/abs/2310.11324
ScreenPoint Medical (2026): MASAI RCT: 12% reduction in interval cancer rate using Transpara. ScreenPoint Medical, Nijmegen, Website, 30. Januar 2026. https://screenpoint-medical.com/insights/final-results-masai-trial (Herstellerangabe) Abgerufen am 25. Oktober 2026.
SECAI – School of Embedded Composite AI (o. J.): SECAI (TU Dresden und Universität Leipzig). Website. https://secai.org/ Abgerufen am 25. Oktober 2026.
Sennrich, Rico; Haddow, Barry; Birch, Alexandra (2016): Neural Machine Translation of Rare Words with Subword Units. ACL 2016, Berlin, S. 1715–1725. https://aclanthology.org/P16-1162/ Abgerufen am 20. Oktober 2026.
Shao, Zhihong u. a. (2024): DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models. Preprint arXiv:2402.03300 (DeepSeek-AI, Tsinghua University, Peking University; zitiert nach v3 vom 27.4.2024), 5. Februar 2024. https://arxiv.org/abs/2402.03300 (Herstellerangabe) Abgerufen am 24. Oktober 2026.
Sharma, B.; Tiwari, A.; Kaur, R. u. a. (2026): The Hidden Thirst of AI: A Framework for Estimating Direct, Indirect, and Scarcity-Adjusted Freshwater Consumption per LLM Query. Green 1 (2), 8 (MDPI); DOI 10.3390/green1020008, 1. September 2026. https://doi.org/10.3390/green1020008 https://www.mdpi.com/3042-9242/1/2/8 Abgerufen am 25. Oktober 2026.
Sharma, Mrinank u. a. (2024): Towards Understanding Sycophancy in Language Models. ICLR 2024; Preprint arXiv:2310.13548 (v1 20.10.2023, v4 10.5.2025), Anthropic. https://arxiv.org/abs/2310.13548 Abgerufen am 20. Oktober 2026.
★ Shazeer, Noam u. a. (2017): Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer. ICLR 2017. https://arxiv.org/abs/1701.06538 Abgerufen am 19. Oktober 2026.
Shehabi, Arman u. a. (2024): 2024 United States Data Center Energy Usage Report. Lawrence Berkeley National Laboratory, LBNL-2001637, Berkeley, Dezember 2024. https://eta-publications.lbl.gov/sites/default/files/2024-12/us_data_center_energy_usage_report_lbnl-2001637_0.pdf Abgerufen am 19. Oktober 2026.
Sheshadri, Abhay u. a. (2025): Why Do Some Language Models Fake Alignment While Others Don’t? Preprint arXiv:2506.18032, 22. Juni 2025. https://arxiv.org/abs/2506.18032 Abgerufen am 25. Oktober 2026.
Shoeybi, Mohammad u. a. (2019): Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism. Preprint, arXiv:1909.08053 (zitiert nach v1), 17. September 2019. https://arxiv.org/abs/1909.08053v1 Abgerufen am 20. Oktober 2026.
Sierra Research (2026): τ³-bench 1.0.0 — Voice, Knowledge, Task Quality (Release Notes). GitHub, Repository tau2-bench, Release v1.0.0, 18. März 2026. https://github.com/sierra-research/tau2-bench/releases/tag/v1.0.0 Abgerufen am 20. Oktober 2026.
Silberling, Amanda (2026): The backlash over OpenAI’s decision to retire GPT-4o shows how dangerous AI companions can be. TechCrunch, 6. Februar 2026. https://techcrunch.com/2026/02/06/the-backlash-over-openais-decision-to-retire-gpt-4o-shows-how-dangerous-ai-companions-can-be/ (Pressebericht) Abgerufen am 25. Oktober 2026.
Silver, David u. a. (2017): Mastering the game of Go without human knowledge. Nature 550, S. 354–359, DOI 10.1038/nature24270, 19. Oktober 2017. https://www.nature.com/articles/nature24270 https://discovery.ucl.ac.uk/id/eprint/10045895/1/agz_unformatted_nature.pdf Abgerufen am 19. Oktober 2026.
Singh, Shivalika u. a. (2025): The Leaderboard Illusion. Preprint. https://arxiv.org/abs/2504.20879
Smith, Lewis; Rajamanoharan, Senthooran; Conmy, Arthur u. a. (2025): Negative Results for SAEs On Downstream Tasks and Deprioritising SAE Research (GDM Mech Interp Team Progress Update #2). AI Alignment Forum (Spiegel des Blogs des DeepMind-Safety-Teams), 26. März 2025. https://www.alignmentforum.org/posts/4uXCAJNuPKtKBsi28/negative-results-for-saes-on-downstream-tasks (Herstellerangabe) Abgerufen am 25. Oktober 2026.
Snell, Charlie; Lee, Jaehoon; Xu, Kelvin u. a. (2024): Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters. Preprint arXiv:2408.03314 (UC Berkeley, Google DeepMind), 6. August 2024. https://arxiv.org/abs/2408.03314 Abgerufen am 20. Oktober 2026.
Sorokovikova, Aleksandra; Chizhov, Pavel; Eremenko, Iuliia u. a. (2025): Surface Fairness, Deep Bias: A Comparative Study of Bias in Language Models. Preprint, arXiv:2506.10491 (v1 12.6.2025, v2 1.9.2025), 12. Juni 2025. https://arxiv.org/abs/2506.10491 Abgerufen am 19. Oktober 2026.
Stability AI (2024): Stable Diffusion 3: Research Paper. Stability AI, News (Blog), 5. März 2024. https://stability.ai/news-updates/stable-diffusion-3-research-paper (Herstellerangabe) Abgerufen am 24. Oktober 2026.
Stanford University (2026): CS224N: Natural Language Processing with Deep Learning (Winter 2026). Kursseite, Stanford. https://web.stanford.edu/class/cs224n/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
★ Strubell, Emma; Ganesh, Ananya; McCallum, Andrew (2019): Energy and Policy Considerations for Deep Learning in NLP. ACL 2019, S. 3645–3650. https://aclanthology.org/P19-1355/ Abgerufen am 19. Oktober 2026.
Struppek, Lukas; McIntyre, Heather; Pelrine, Kellin (2026): Security Stress Test: Exposing the Brittleness of DeepSeek-V4-Pro’s Safeguards. FAR.AI, Blog, 11. Mai 2026. https://www.far.ai/blog/security-stress-test-deepseek-v4-pros-safeguards Abgerufen am 20. Oktober 2026.
Su, Jianlin; Lu, Yu; Pan, Shengfeng u. a. (2021): RoFormer: Enhanced Transformer with Rotary Position Embedding. Preprint arXiv:2104.09864 (v1 vom 20.4.2021, v5 vom 8.11.2023); Zeitschriftenfassung: Neurocomputing 568 (Februar 2024), Art. 127063, DOI 10.1016/j.neucom.2023.127063, dort in der Reihenfolge Su, Ahmed, Lu, Pan, Bo, Liu, 20. April 2021. https://arxiv.org/abs/2104.09864 https://doi.org/10.1016/j.neucom.2023.127063 Abgerufen am 23. Oktober 2026.
Talus, Anu (2026): US Supreme Court judgment Trump v. Slaughter. Brief des European Data Protection Board an Kommissar Michael McGrath, Brüssel, 31. Juli 2026. https://www.edpb.europa.eu/system/files/2026-08/edpb_letter_20260731_us_supremecourt_judgment_trump_v_slaughter_en.pdf Abgerufen am 19. Oktober 2026.
Tamirisa, Rishub u. a. (2024): Tamper-Resistant Safeguards for Open-Weight LLMs. Preprint arXiv:2408.00761, 1. August 2024. https://arxiv.org/abs/2408.00761 Abgerufen am 25. Oktober 2026.
Team Olmo (2025): Olmo 3. Preprint arXiv:2512.13961 (Allen Institute for AI; v2 vom 14.4.2026, zitierte Stellen unverändert), 15. Dezember 2025. https://arxiv.org/abs/2512.13961 (Herstellerangabe) Abgerufen am 24. Oktober 2026.
TelefonSeelsorge Deutschland (2026): TelefonSeelsorge Deutschland: Sorgen kann man teilen (Startseite mit Rufnummern). Website. https://www.telefonseelsorge.de/ Abgerufen am 2. November 2026.
Templeton, Adly u. a. (2024): Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet. Transformer Circuits Thread (Anthropic), nicht begutachtet, 21. Mai 2024. https://transformer-circuits.pub/2024/scaling-monosemanticity/ (Herstellerangabe) Abgerufen am 19. Oktober 2026.
Terminal-Bench (2026a): Terminal-Bench 3.0. Website, 30. Juli 2026. https://www.tbench.ai/news/terminal-bench-3-0 Abgerufen am 20. Oktober 2026.
Terminal-Bench (2026b): Terminal-Bench 4.0. Website, 28. August 2026. https://www.tbench.ai/news/terminal-bench-4-0
Terminal-Bench (o. J.): Terminal-Bench (Code und Aufgabenformat). GitHub. https://github.com/harbor-framework/terminal-bench
The Hill (2025): Character.AI to ban children from talking with chatbots. The Hill, Technology, 29. Oktober 2025. https://thehill.com/policy/technology/5578891-character-ai-children-chatbots-suicide (Pressebericht) Abgerufen am 25. Oktober 2026.
The New York Times Company (2023): Complaint, The New York Times Company v. Microsoft Corporation, OpenAI, Inc. et al. Case 1:23-cv-11195, U.S. District Court for the Southern District of New York (Klageschrift), 27. Dezember 2023. https://nytco-assets.nytimes.com/2023/12/NYT_Complaint_Dec2023.pdf Abgerufen am 19. Oktober 2026.
The Nobel Committee for Physics (2024): Scientific Background to the Nobel Prize in Physics 2024: „For foundational discoveries and inventions that enable machine learning with artificial neural networks“. Königlich Schwedische Akademie der Wissenschaften, Stockholm, 8. Oktober 2024. https://www.nobelprize.org/uploads/2024/11/advanced-physicsprize2024-3.pdf (Sekundärquelle) Abgerufen am 20. Oktober 2026.
The Royal Swedish Academy of Sciences (2024a): Popular information: They have revealed proteins’ secrets through computing and artificial intelligence (Nobel Prize in Chemistry 2024). Königlich Schwedische Akademie der Wissenschaften, Stockholm (nobelprize.org), 9. Oktober 2024. https://www.nobelprize.org/prizes/chemistry/2024/popular-information/ Abgerufen am 25. Oktober 2026.
The Royal Swedish Academy of Sciences (2024b): Press release: The Nobel Prize in Chemistry 2024. Königlich Schwedische Akademie der Wissenschaften, Stockholm (nobelprize.org), 9. Oktober 2024. https://www.nobelprize.org/prizes/chemistry/2024/press-release/ Abgerufen am 25. Oktober 2026.
Thinking Machines Lab (o. J.): Tinker. Thinking Machines Lab, Produktseite mit FAQ „What is LoRA and why does Tinker use it?“ (ohne Datum). https://thinkingmachines.ai/tinker/ (Herstellerangabe) Abgerufen am 24. Oktober 2026.
Tian, Katherine u. a. (2023): Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback. EMNLP 2023; Preprint arXiv:2305.14975, 24. Mai 2023. https://arxiv.org/abs/2305.14975 Abgerufen am 20. Oktober 2026.
TIB – Leibniz-Informationszentrum Technik und Naturwissenschaften (2024): Neuer Dienst „ORKG Ask“ gestartet; dazu About ORKG Ask und Statistics. Pressemitteilung (16.5.2024) und Website (Stand 2.11.2026), Hannover, 16. Mai 2024. https://nachrichten.idw-online.de/2024/05/16/neuer-dienst-orkg-ask-gestartet-antworten-aus-80-millionen-wissenschaftlichen-publikationen https://ask.orkg.org/pages/about https://ask.orkg.org/stats Abgerufen am 2. November 2026.
TIGER-AI-Lab (o. J.): MMLU-Pro (Code und Auswertungsskripte). GitHub. https://github.com/TIGER-AI-Lab/MMLU-Pro
TIGER-Lab (o. J.): MMLU-Pro (Datensatz und Beschreibung). Hugging Face. https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro
Touvron, Hugo u. a. (2023): LLaMA: Open and Efficient Foundation Language Models. Preprint, arXiv:2302.13971v1 (Meta AI), 27. Februar 2023. https://arxiv.org/abs/2302.13971 (Herstellerangabe) Abgerufen am 24. Oktober 2026.
TrendForce (2026a): Memory Price Outlook for 1Q26 Sharply Upgraded; QoQ Increases of All Product Categories to Hit Record Highs, Says TrendForce. TrendForce, Pressemitteilung, 2. Februar 2026. https://www.trendforce.com/presscenter/news/20260202-12911.html (Sekundärquelle) Abgerufen am 19. Oktober 2026.
TrendForce (2026b): AI Server Demand Sustains Memory Contract Price Increases in 4Q26, While Consumer-Side Pressure Persists, Says TrendForce. TrendForce, Pressemitteilung, 30. September 2026. https://www.trendforce.com/presscenter/news/20260930-13258.html (Sekundärquelle) Abgerufen am 2. November 2026.
TSMC (o. J.): CoWoS. TSMC 3DFabric, Technologieseite. https://3dfabric.tsmc.com/english/dedicatedFoundry/technology/cowos.htm (Herstellerangabe) Abgerufen am 19. Oktober 2026.
TU Dortmund, Lehrstuhl Wissenschaftsjournalismus (o. J.): Wissenschaftsjournalismus an der TU Dortmund. Website des Lehrstuhls. https://www.wissenschaftsjournalismus.org/ https://wissenschaftsjournalismus.tu-dortmund.de/ Abgerufen am 25. Oktober 2026.
Turpin, Miles; Michael, Julian; Perez, Ethan u. a. (2023): Language Models Don’t Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting. NeurIPS 2023; Preprint arXiv:2305.04388 (NYU, Cohere, Anthropic), 7. Mai 2023. https://arxiv.org/abs/2305.04388 Abgerufen am 20. Oktober 2026.
Umweltbundesamt (2026a): CO₂-Emissionen pro Kilowattstunde Strom 2025 nur leicht gesunken. Dessau-Roßlau, 23. März 2026. https://www.umweltbundesamt.de/themen/co2-emissionen-pro-kilowattstunde-strom-2025-nur Abgerufen am 19. Oktober 2026.
Umweltbundesamt (2026b): Rechenzentrumsregister: Daten widerlegen hohe Rechtsanforderungen. Dessau-Roßlau, 21. September 2026. https://www.umweltbundesamt.de/themen/rechenzentrumsregister-daten-widerlegen-hohe Abgerufen am 19. Oktober 2026.
United States District Court for the Northern District of California (2025): Bartz v. Anthropic PBC, Order on Fair Use. Nr. 3:24-cv-05417-WHA, Dok. 231, 23. Juni 2025. https://storage.courtlistener.com/recap/gov.uscourts.cand.434709/gov.uscourts.cand.434709.231.0_3.pdf Abgerufen am 20. Oktober 2026.
University of Helsinki; MinnaLearn (o. J.): Elements of AI (deutsche Fassung). Kursseite (ohne Datum). https://www.elementsofai.de/ (Sekundärquelle) Abgerufen am 25. Oktober 2026.
U.S. Copyright Office (2025): Copyright and Artificial Intelligence, Part 2: Copyrightability. Washington, D.C., 29. Januar 2025. https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf Abgerufen am 20. Oktober 2026.
U.S. Food and Drug Administration (2026): Artificial Intelligence-Enabled Medical Devices (Liste, laufend aktualisiert; neuester Eintrag 29.6.2026). FDA, Silver Spring (Maryland). https://www.fda.gov/medical-devices/artificial-intelligence-enabled-medical-devices/list-artificial-intelligence-enabled-medical-devices Abgerufen am 25. Oktober 2026.
Valve (2026): Steam Hardware & Software Survey: October 2026. store.steampowered.com, November 2026. https://store.steampowered.com/hwsurvey/Steam-Hardware-Software-Survey-Welcome-to-Steam Abgerufen am 2. November 2026.
van den Oord, Aaron; Dieleman, Sander; Zen, Heiga u. a. (2016): WaveNet: A Generative Model for Raw Audio. Preprint arXiv:1609.03499 (DeepMind; v2 19.9.2016), 12. September 2016. https://arxiv.org/abs/1609.03499 Abgerufen am 25. Oktober 2026.
★ Vaswani, Ashish u. a. (2017): Attention Is All You Need. NIPS 2017; arXiv-Fassungen v1 (12.6.2017) bis v7 (2.8.2023). https://arxiv.org/abs/1706.03762 Abgerufen am 19. Oktober 2026.
Verbraucherzentrale Brandenburg (2026): KI and me. Schockanrufe in neuem Gewand. Pressemitteilung, Potsdam, 26. Januar 2026. https://www.verbraucherzentrale-brandenburg.de/pressemitteilungen/ki-and-me-schockanrufe-in-neuem-gewand-116706 Abgerufen am 25. Oktober 2026.
Verbraucherzentrale NRW (2025): OLG Köln weist Eilantrag der Verbraucherzentrale NRW gegen Meta wegen KI-Training zurück. Urteilsseite der Klägerin zu OLG Köln, Urt. v. 23.5.2025, 15 UKl 2/25, mit Urteils-PDF, Stand 18.6.2025, Düsseldorf. https://www.verbraucherzentrale.nrw/urteile/olg-koeln-weist-eilantrag-der-verbraucherzentrale-nrw-gegen-meta-wegen-kitraining-zurueck-108065 Abgerufen am 19. Oktober 2026.
Vereinigte Staaten (2018): 18 U.S.C. § 2713: Required preservation and disclosure of communications and records (CLOUD Act, Pub. L. 115–141, div. V, § 103(a)(1)). Legal Information Institute, Cornell, 23. März 2018. https://www.law.cornell.edu/uscode/text/18/2713 Abgerufen am 19. Oktober 2026.
Verma, Pranshu; Tan, Shelly (2024): A bottle of water per email: the hidden environmental costs of using AI chatbots. The Washington Post, 18. September 2024. https://www.washingtonpost.com/technology/2024/09/18/energy-ai-use-electricity-water-data-centers/ (Pressebericht) Abgerufen am 19. Oktober 2026.
Virginia Tech (2009): In Memoriam: I. J. Good, University Distinguished Professor and pioneer of modern statistics. Virginia Tech News, Blacksburg, 6. April 2009. https://news.vt.edu/articles/2009/04/2009-276.html Abgerufen am 2. November 2026.
Wang, Xuezhi u. a. (2022): Self-Consistency Improves Chain of Thought Reasoning in Language Models. ICLR 2023; Preprint arXiv:2203.11171 (Google Research), 21. März 2022. https://arxiv.org/abs/2203.11171 Abgerufen am 20. Oktober 2026.
WEB.DE (2025): Einen halben Liter pro Anfrage: Warum KI so viel Wasser verbraucht. WEB.DE Magazin, Wissen (Video, Bildnachweis epd; aktualisiert am 22.5.2025), 22. Mai 2025. https://web.de/magazine/wissen/videos/wissenschaft-technik/halben-liter-anfrage-ki-wasser-verbraucht-41004366 (Pressebericht) Abgerufen am 19. Oktober 2026.
★ Wei, Alexander; Haghtalab, Nika; Steinhardt, Jacob (2023): Jailbroken: How Does LLM Safety Training Fail? NeurIPS 2023. https://proceedings.neurips.cc/paper_files/paper/2023/hash/fd6613131889a4b656206c50a8bd7790-Abstract-Conference.html Abgerufen am 19. Oktober 2026.
★ Wei, Jason u. a. (2022): Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. NeurIPS 2022. https://arxiv.org/abs/2201.11903 Abgerufen am 19. Oktober 2026.
Wei, Jason u. a. (2024): Measuring short-form factuality in large language models. Preprint. https://arxiv.org/abs/2411.04368
Weiß, Eva-Maria (2024): ChatGPTs Stromverbrauch: Zehnmal mehr als bei Google. heise online, 30. August 2024. https://www.heise.de/news/ChatGPTs-Stromverbrauch-Zehnmal-mehr-als-bei-Google-9852126.html (Pressebericht) Abgerufen am 19. Oktober 2026.
Weizenbaum, Joseph (1966): ELIZA – A Computer Program for the Study of Natural Language Communication between Man and Machine. Communications of the ACM 9 (1), S. 36–45, DOI 10.1145/365153.365168, Januar 1966. https://doi.org/10.1145/365153.365168 https://web.stanford.edu/class/cs124/p36-weizenabaum.pdf Abgerufen am 19. Oktober 2026.
Weizenbaum-Institut (2025): Zum 103. Geburtstag von Joseph Weizenbaum – Ein Visionär der digitalen Gesellschaft. Nachrichtenbeitrag, Berlin, 19. Dezember 2025. https://www.weizenbaum-institut.de/news/detail/zum-103-geburtstag-von-joseph-weizenbaum-ein-visionaer-der-digitalen-gesellschaft (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Wiggers, Kyle (2025a): OpenAI adopts rival Anthropic’s standard for connecting AI models to data. TechCrunch, 26. März 2025. https://techcrunch.com/2025/03/26/openai-adopts-rival-anthropics-standard-for-connecting-ai-models-to-data/ (Pressebericht) Abgerufen am 19. Oktober 2026.
Wiggers, Kyle (2025b): Google to embrace Anthropic’s standard for connecting AI models to data. TechCrunch, 9. April 2025. https://techcrunch.com/2025/04/09/google-says-itll-embrace-anthropics-standard-for-connecting-ai-models-to-data/ (Pressebericht) Abgerufen am 19. Oktober 2026.
Wiggers, Kyle (2025c): Meta’s vanilla Maverick AI model ranks below rivals on a popular chat benchmark. TechCrunch, 11. April 2025. https://techcrunch.com/2025/04/11/metas-vanilla-maverick-ai-model-ranks-below-rivals-on-a-popular-chat-benchmark/ (Pressebericht)
Wijk, Hjalmar; Cotra, Ajeya; Greenblatt, Ryan (2026): Hugging Face incident investigation report (Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident). METR und Redwood Research, PDF, 26. August 2026. https://metr.org/hugging-face-incident-report-aug-2026.pdf Abgerufen am 25. Oktober 2026.
Willison, Simon (2022): Prompt injection attacks against GPT-3. Blog simonwillison.net, 12. September 2022. https://simonwillison.net/2022/Sep/12/prompt-injection/ Abgerufen am 19. Oktober 2026.
Willison, Simon (2025a): The lethal trifecta for AI agents: private data, untrusted content, and external communication. Blog simonwillison.net, 16. Juni 2025. https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/ Abgerufen am 19. Oktober 2026.
Willison, Simon (2025b): A quote from Workaccount2 on Hacker News. Blog simonwillison.net (zitiert den Hacker-News-Kommentar des Nutzers „Workaccount2“ vom 18.6.2025, Beitrag 44310054, mit dem Begriff „context rot“), 18. Juni 2025. https://simonwillison.net/2025/Jun/18/context-rot/ https://news.ycombinator.com/item?id=44308711#44310054 (Sekundärquelle) Abgerufen am 23. Oktober 2026.
Willison, Simon (2025c): Context engineering. Blog simonwillison.net, 27. Juni 2025. https://simonwillison.net/2025/Jun/27/context-engineering/ (Sekundärquelle) Abgerufen am 23. Oktober 2026.
Willison, Simon (2025d): Our contribution to a global environmental standard for AI. Blog simonwillison.net (Blogeintrag mit dem Text der Mistral-Infografik), 22. Juli 2025. https://simonwillison.net/2025/Jul/22/mistral-environmental-standard/ (Sekundärquelle) Abgerufen am 20. Oktober 2026.
Wilson, Edwin B. (1927): Probable Inference, the Law of Succession, and Statistical Inference. Journal of the American Statistical Association 22 (158): 209–212. https://doi.org/10.1080/01621459.1927.10502953 Abgerufen am 19. Oktober 2026.
Wohlrabe, Klaus (2025): Mehr KI, weniger Jobs? Was Unternehmen in Deutschland erwarten. ifo Schnelldienst digital 6 (8), München, 16. Juni 2025. https://www.ifo.de/publikationen/2025/aufsatz-zeitschrift/mehr-ki-weniger-jobs-was-unternehmen-deutschland-erwarten Abgerufen am 20. Oktober 2026.
World Resources Institute; World Business Council for Sustainable Development (2004): The Greenhouse Gas Protocol – A Corporate Accounting and Reporting Standard, Revised Edition. Greenhouse Gas Protocol, März 2004. https://ghgprotocol.org/sites/default/files/standards/ghg-protocol-revised.pdf Abgerufen am 19. Oktober 2026.
World Resources Institute; World Business Council for Sustainable Development (2015): GHG Protocol Scope 2 Guidance. Greenhouse Gas Protocol, Januar 2015. https://ghgprotocol.org/sites/default/files/2023-03/Scope%202%20Guidance.pdf Abgerufen am 19. Oktober 2026.
Xie, Tianbao u. a. (2024): OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments. NeurIPS 2024. https://arxiv.org/abs/2404.07972
Xu, Shusheng u. a. (2024): Is DPO Superior to PPO for LLM Alignment? A Comprehensive Study. ICML 2024; Preprint arXiv:2404.10719 (Tsinghua University, OpenPsi Inc., Shanghai Qi Zhi Institute). https://arxiv.org/abs/2404.10719 Abgerufen am 25. Oktober 2026.
Xu, Zhi-Qin John; Luo, Tao; Ma, Zheng u. a. (2020): MLMath – Mathematical notation for Machine Learning (Suggested Notation for Machine Learning). BAAI, Peking University, Shanghai Jiao Tong University; CTAN-Paket mlmath 1.0.0, 18. Juni 2020. https://ctan.org/pkg/mlmath Abgerufen am 19. Oktober 2026.
Xuan, Weihao u. a. (2025): MMLU-ProX: A Multilingual Benchmark for Advanced Large Language Model Evaluation. Preprint, arXiv:2503.10497 (letzte Fassung 26.5.2025), 13. März 2025. https://arxiv.org/abs/2503.10497 Abgerufen am 20. Oktober 2026.
Yao, Shunyu u. a. (2024): τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains. ICLR 2025. https://arxiv.org/abs/2406.12045
Yin, Leon; Alba, Davey; Nicoletti, Leonardo (2024): OpenAI’s GPT Is a Recruiter’s Dream Tool. Tests Show There’s Racial Bias. Bloomberg, Technology + Equality, 7. März 2024. https://www.bloomberg.com/graphics/2024-openai-gpt-hiring-racial-discrimination/ (Pressebericht) Abgerufen am 25. Oktober 2026.
Yuan, Mengqi u. a. (2026): OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks. Preprint, arXiv:2606.29537, 28. Juni 2026. https://arxiv.org/abs/2606.29537 https://osworld-v2.xlang.ai/
Yue, Yang u. a. (2025): Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model? NeurIPS 2025 (Best Paper Runner-up); Preprint arXiv:2504.13837 (Tsinghua University, Shanghai Jiao Tong University). https://arxiv.org/abs/2504.13837 Abgerufen am 25. Oktober 2026.
Zeff, Maxwell (2025): DeepSeek displaces ChatGPT as the App Store’s top app. TechCrunch, 27. Januar 2025. https://techcrunch.com/2025/01/27/deepseek-displaces-chatgpt-as-the-app-stores-top-app/ (Pressebericht) Abgerufen am 26. Oktober 2026.
Zheng, Lianmin u. a. (2023): Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena. NeurIPS 2023, Datasets and Benchmarks Track. https://arxiv.org/abs/2306.05685
Zhong, Yinmin u. a. (2024): DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving. OSDI 2024; Preprint arXiv:2401.09670 (8 Autoren), 18. Januar 2024. https://arxiv.org/abs/2401.09670 Abgerufen am 19. Oktober 2026.
Zhou, Jerry; Chen, Richard (2025): NVIDIA DGX Spark In-Depth Review: A New Standard for Local AI Inference. LMSYS Blog, 13. Oktober 2025. https://lmsys.org/blog/2025-10-13-nvidia-dgx-spark/ (Sekundärquelle) Abgerufen am 19. Oktober 2026.
Zhou, Kaiwen; Liu, Chengzhi; Zhao, Xuandong u. a. (2025): The Hidden Risks of Large Reasoning Models: A Safety Assessment of R1. Preprint, arXiv:2502.12659 (zitiert nach v4 vom 17.11.2025), 18. Februar 2025. https://arxiv.org/abs/2502.12659 Abgerufen am 20. Oktober 2026.