Poster · Sydney
SFT-then-RL Outperforms Mixed-Policy Methods for LLM Reasoning
- When
- Tue 8 Dec 2026, 10:00–13:00 AEDT
- Where
- Hall 1-4, ICC Sydney
Exhibition Centre, Level 1 - Session
- Sydney Poster Session 1
- Authors
- Alexis Limozin, Eduard F Ďurech, Torsten Hoefler, Imanol Schlag, Valentina Pyatkin
- Organisations
- ETH Zurich, EPFL
- QRS classification
- Reinforcement learning, Language models