← Search

Lukas Berglund

2 accepted papers

2025

AssistanceZero: Scalably Solving Assistance Games

ICML 2025poster

Assistance games are a promising alternative to reinforcement learning from human feedback (RLHF) for training AI assistants. Assistance games resolve key drawbacks of RLHF, such as incentives for deceptive behavior, by explicitly modeling the interaction between assistant and user as a two-player g…

2024

The Reversal Curse: LLMs trained on “A is B” fail to learn “B is A”

ICLR 2024poster

We expose a surprising failure of generalization in auto-regressive large language models (LLMs). If a model is trained on a sentence of the form ''_A_ is _B_'', it will not automatically generalize to the reverse direction ''_B_ is _A_''. This is the **Reversal Curse**. For instance, if a model is…