Poster · Sydney
STARE: Surprisal-Guided Token-Level Advantage Reweighting for Policy Entropy Stability
- When
- Thu 10 Dec 2026, 10:00–13:00 AEDT
- Where
- Hall 1-4, ICC Sydney
Exhibition Centre, Level 1 - Session
- Sydney Poster Session 5
- Authors
- HAIPENG LUO, Qingfeng Sun, Song-Li Wu, Can Xu, Wenfeng Deng, Han Hu, Yansong Tang
- Organisations
- Tsinghua University, Microsoft, Pengcheng Laboratory, Microsft Research Asia
- QRS classification
- Reinforcement learning