The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning
Paper • 2606.29526 • Published • 137
Hongyao Tang
thy-1now
·
AI & ML interests
Reinforcement Learning, Representation Learning
Recent Activity
upvoted a paper 3 days ago
Trust-Region Behavior Blending for On-Policy Distillation upvoted a paper 22 days ago
Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation ModelsOrganizations
None yet