Learning Spatiotemporal Sensitivity in Video LLMs via Counterfactual Reinforcement Learning
Paper • 2605.21988 • Published • 1
🔄 In a Training Loop
dudazhao
ddz16
AI & ML interests
None yet
Recent Activity
authored a paper about 1 hour ago
Temporally Grounded Compositional Camera Motion Understanding via Geometric Knowledge Distillation authored a paper about 1 hour ago
WorldCycle: Self-Verifiable Reinforcement Learning for Long-Horizon Video World Models authored a paper about 1 hour ago
WorldCraft: From Camera Navigation to Object Manipulation in Interactive Video World ModelsOrganizations
None yet
upvoted a paper about 1 month ago