AI Safety TübingenEnglish
Zurück zur Startseite

Archiv

Paper, die wir diskutiert haben, Vorträge, die wir organisiert haben, und Socials, die wir genossen haben.

Vergangene Events

  1. Social

    Summer social at Neckawa!

  2. Paper-Diskussion

    AI Welfare: Is it bullshit or should we take it seriously?

    Jeanne Salle

  3. Vortrag

    AI Control: Overview and ResearchArena paper

    Lena Libon

  4. Social

    AI Safety Pubquiz

  5. Vortrag

    Models That Know How Evaluations Are Designed Score Safer

    Katharina Deckenbach

  6. Paper-Diskussion

    Europe2031 – What getting AI wrong means for us

  7. Vortrag

    Unlearning as an Asymmetric Generalization Problem

    Amit Peleg

  8. Social

    Watchparty – The AI Doc: Or how I became an Apocaloptimist

  9. Event

    MATS Q&A with three recent MATS alumni

  10. Vortrag

    Claudini: Autoresearch Discovers State-of-the-Art Adversarial Attack Algorithms for LLMs

    Alexander Panfilov

  11. Social

    AI Calibration Game

  12. Paper-Diskussion

    Claude Mythos System Card

  13. Vortrag

    Compute Governance

    Yannick Mühlhäuser (FLI)

  14. Paper-Diskussion

    Legal Alignment for Safe and Ethical AI

    Kolt et al. (2026)

    arXiv

  15. Paper-Diskussion

    The Artificial Self: Characterising the landscape of AI identity

    Douglas et al. (2026)

  16. Vortrag

    AI Personas

    Daniel Tan

  17. Paper-Diskussion

    Who’s in Charge? Disempowerment Patterns in Real-World LLM Usage

    Sharma et al. (2026)

    arXiv 2510.04340 arXiv 2601.19062

  18. Vortrag

    From Machine Psychology to AI Safety: Studying Deception, Emergent Misalignment, and Self-Awareness in LLMs

    Thilo Hagendorff (Group Leader at University of Stuttgart)

  19. Social

    Speed-Friending & What are you working on?

  20. Social

    Casual Dinner @ Irish Pub Tübingen

  21. Vortrag

    SKILL-INJECT: Measuring Agent Vulnerability to Skill File Attacks

    David Schmotz

  22. Vortrag

    Measuring and Forecasting Autonomous Capabilities

    Jeanne Salle

    METR blog post

  23. Paper-Diskussion

    Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment

    alignmentpretraining.ai

  24. Paper-Diskussion

    Spilling the beans – Teaching LLMs to Self-Report Their Hidden Objectives, and its predecessor: Teaching Models to Verbalize Reward Hacking in Chain-of-Thought Reasoning

    arXiv 2511.06626 arXiv 2506.22777

  25. Paper-Diskussion

    Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers

    Anthropic

  26. Vortrag

    Intro to Scalable Oversight

    Ameya Prabhu (Bethgelab)

  27. Paper-Diskussion

    Quick introduction + Natural Emergent Misalignment in Production RL (Anthropic)

    Anthropic