Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
Paper • 2609.18708 • Published • 79
Kanghui Tian
KangHui2200
·
AI & ML interests
None yet
Recent Activity
upvoted a paper 8 days ago
Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening upvoted a paper 4 months ago
InternVideo3: Agentify Foundation Models with Multimodal Contextual Reasoning upvoted a paper 4 months ago
KVPO: ODE-Native GRPO for Autoregressive Video Alignment via KV Semantic ExplorationOrganizations
None yet