AI Paper Atlas

Research map / Video / Video-language understanding

ReMoT: Reinforcement Learning with Motion Contrast Triplets

Cong Wan, Zeyu Guo, Jiangyang Li, Songlin Dong, Yifan Bai, Lin Peng, Zhiheng Ma, Yihong Gong

CVPR 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add alternating SFT and GRPO training instead of pure GRPO (ablation)Qwen3-VL-4B-Thinking · ReMoT-16k-Test (avg, overall)overall accuracy (%): 33.6 -> 38.0Improved
add GRPO training on motion-contrast data to the base model (ablation)Qwen3-VL-4B-CoT · ReMoT-16k-Test (avg, overall)overall accuracy (%): 20.7 -> 33.6Improved
add alternating SFT instead of sequential SFT then GRPO (ablation)Qwen3-VL-4B-Thinking · ReMoT-16k-Test Camera (navigation, overall)overall accuracy (%): 27.0 -> 21.4Hurt
add navigation data to manipulation-only training (ablation)ReMoT training · ReMoT-16k-Test (overall)overall accuracy (%): 23.9 -> 32.4Improved
add simulation data (ablation)ReMoT training · ReMoT-16k-Test (overall)overall accuracy (%): 32.4 -> 38.0Improved
add logic consistency reward to GRPO (ablation)Qwen3-VL-4B-CoT · Manipulation subset (logic test)overall accuracy (%): 68.6 -> 78.0Improved
add logic consistency reward to GRPO (ablation)Qwen3-VL-4B-CoT · Manipulation subset (logic test)logic consistency (%): 98.6 -> 99.3Improved
build training data with multi-expert rule-based pipeline instead of VLM-generated data (ablation)ReMoT data construction · ReMoT-16k-Test scalingaccuracy: 0.49 -> 0.66Improved
apply ReMoT training to Qwen3-VL-4B on spatial-temporal benchmarksQwen3-VL-4B-CoT · VLM2accuracy (%): 66.4 -> 70.0Improved
apply ReMoT training to Qwen3-VL-4B on general benchmarksQwen3-VL-4B-CoT · MUIRaccuracy (%): 73.8 -> 72.8Hurt
apply ReMoT training to Qwen3-VL-4B on general benchmarksQwen3-VL-4B-CoT · MMMUaccuracy (%): 70.8 -> 71.4No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy