List your company

Video Reasoning on SEED-Bench-R1 (L1 In-Dist.)

50.5AccuracyAPPO
27.30833.32939.3545.371Feb 27, 2026

Evaluation results

MethodLinks
APPOBackbone=Qwen2.5-VL-7B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0250.5
DAPOBackbone=Qwen2.5-VL-7B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0250
GRPOBackbone=Qwen2.5-VL-7B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0249
SFTBackbone=Qwen2.5-VL-7B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0240.2
APPOBackbone=Qwen2.5-VL-3B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0237.5
DAPOBackbone=Qwen2.5-VL-3B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0236.6
APPOSize=7B, Training Data=34K, Resolution=224 x 224, Sampling Rate (fps)=1fps, Maximum Frames=30, Zero-shot evaluation=true2026.0235.4
GRPOBackbone=Qwen2.5-VL-3B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0235.3
VideoChat-R1Size=7B, Training Data=18K, Resolution=224 x 224, Sampling Rate (fps)=1fps, Maximum Frames=30, Zero-shot evaluation=true2026.0233.3
SFTBackbone=Qwen2.5-VL-3B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0232.6
VideoRFTSize=7B, Training Data=310K, Resolution=224 x 224, Sampling Rate (fps)=1fps, Maximum Frames=30, Zero-shot evaluation=true2026.0232.4
Video-R1Size=7B, Training Data=260K, Resolution=224 x 224, Sampling Rate (fps)=1fps, Maximum Frames=30, Zero-shot evaluation=true2026.0230.9
TW-GRPOSize=7B, Training Data=1K, Resolution=224 x 224, Sampling Rate (fps)=1fps, Maximum Frames=30, Zero-shot evaluation=true2026.0230.2
GRPO-CARESize=7B, Training Data=260K, Resolution=224 x 224, Sampling Rate (fps)=1fps, Maximum Frames=30, Zero-shot evaluation=true2026.0229.9
Base ModelBackbone=Qwen2.5-VL-7B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0229.1
Base ModelBackbone=Qwen2.5-VL-3B, Resolution=224x224, Sampling Rate=1 fps, Max Frames=602026.0228.2