Poster · Atlanta

Mode-Controlled Policy Optimization: A Geometry-Aware Recipe for LLM Post-Training

Name: Mode-Controlled Policy Optimization: A Geometry-Aware Recipe for LLM Post-Training
Start: 2026-12-09T16:30:00-05:00
End: 2026-12-09T19:30:00-05:00
Location: Hall C1, Georgia World Congress Center

When: Wed 9 Dec 2026, 16:30–19:30 EST
Where: Hall C1, Georgia World Congress Center
Session: Atlanta Poster Session 2
Authors: Zhenyu Sun, Dylan Hadfield-Menell, Weiguo Feng, Xiaohuan Zhou, Yi Zeng
Organisations: ByteDance, Massachusetts Institute of Technology
QRS classification: Reinforcement learning, Agentic AI, Language models

Open in schedule Official page