List your company

Video Reasoning on MMVU Multiple-choice (test)

75.4AccuracyGPT-4o
43.57651.83860.168.362Nov 25, 2025Dec 19, 2025Jan 13, 2026Feb 7, 2026Mar 3, 2026Mar 28, 2026Apr 22, 2026

Evaluation results

MethodLinks
GPT-4oFrames=-2026.0475.4
GPT-4oFrame Count=16-frame, Activation Replay=false2025.1175.4
Video-ToCFrames=642026.0466.5
Video-ToCFrames=162026.0466.1
Video-ToCFrames=322026.0466.1
VideoChat-R1Frames=162026.0465.8
Video-ToC-SFTFrames=322026.0465.4
Video-ToC-SFTFrames=642026.0465.4
Video-ToC-SFTFrames=162026.0465.3
Video-R1-7B + replayFrame Count=16-frame, Activation Replay=true2025.1165
Video-R1Frames=162026.0464.2
Video-R1-7BFrame Count=16-frame, Activation Replay=false2025.1164.2
Video-R1Frames=322026.0463.8
Video-R1Frames=642026.0463.8
Video-R1-SFTFrames=322026.0463.5
Video-R1-SFTFrames=642026.0461.6
Video-R1-SFTFrames=162026.0461.3
Baseline (Qwen2.5-VL-7B)Frames=322026.0460
Baseline (Qwen2.5-VL-7B)Frames=642026.0460
Baseline (Qwen2.5-VL-7B)Frames=162026.0459.2
Qwen2.5-VL-7BFrame Count=16-frame, Activation Replay=false2025.1150.1
LLaVA-OneVision-7BFrames=-2026.0449.2
LLaVA-OV-7BFrame Count=16-frame, Activation Replay=false2025.1149.2
TinyLLaVA-Video-R1Frames=162026.0446.9
VideoLLaMA2-7BFrames=-2026.0444.8