Jakob Foerster

发表

Rethinking Out-of-Distribution Detection for Reinforcement Learning: Advancing Methods for Evaluation and Detection

C. S. D. Witt, L. Nasvytis, Tim Franzmeyer, 2024, AAMAS.

Illusory Attacks: Information-Theoretic Detectability Matters in Adversarial Attacks

C. S. D. Witt, Tim Franzmeyer, P. Torr, 2022, 2207.10170.

Risks and Opportunities of Open-Source Generative AI

Joseph Marvin Imperial, Bertie Vidgen, Fazel Keshtkar, 2024, ArXiv.