2 papers
cs.CL2024
Scalable Reinforcement Post-Training Beyond Static Human Prompts: Evolving Alignment via Asymmetric Self-Play
Ziyu Ye, Rishabh Agarwal, Tianqi Liu +5
Current reinforcement learning (RL) frameworks for large language models (LLM) post-training typically assume a fixed prompt distribution, which is sub-optimal and bottlenecks scal…
cs.LG2024
SiT: Symmetry-Invariant Transformers for Generalisation in Reinforcement Learning
Matthias Weissenbacher, Rishabh Agarwal, Yoshinobu Kawahara
An open challenge in reinforcement learning (RL) is the effective deployment of a trained policy to new or slightly different situations as well as semantically-similar environment…