Video-Based Optimal Transport for Feedback-Efficient Offline Preference-Based Reinforcement Learning (ICML 2026 Oral)
-
Updated
Sep 18, 2026 - Python
Video-Based Optimal Transport for Feedback-Efficient Offline Preference-Based Reinforcement Learning (ICML 2026 Oral)
Frozen V-JEPA 2 (300M params) + PPO policy head achieves +763 mean reward on CarRacing-v3 — 2.4x over CNN baseline, single RTX 3080
Investigating whether video foundation models encode intuitive physics, through layerwise probing of their frozen representations.
Event detection and zero-shot spatial localization with V-JEPA 2.1, PEFT, Integrated Grandients and VTCD
To associate your repository with the video-foundation-models topic, visit your repo's landing page and select "manage topics."