Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Social Reasoning on TVQA+
Loading...
73.2
Accuracy
Qwen3-VL-8B + SGR
58.848
62.574
66.3
70.026
May 15, 2026
Accuracy
Updated 16d ago
Evaluation Results
Method
Method
Links
Accuracy
Qwen3-VL-8B + SGR
Training Type=Socially...
2026.05
73.2
LLaVA-OV-1.5-8B-Instruct
Training Type=Supervis...
2026.05
72.5
Qwen3-VL-8B-Instruct
Training Type=Supervis...
2026.05
71.9
Video-R1-7B
Training Type=RL Post-...
2026.05
71.2
InternVL3.5-8B
Training Type=RL Post-...
2026.05
70.9
Qwen3.5-9B + SGR
Training Type=Socially...
2026.05
70.9
Qwen3.5-9B (Instruct Mode)
Training Type=Supervis...
2026.05
70.6
Qwen2.5-VL-7B-Instruct
Training Type=Supervis...
2026.05
70.4
Qwen3-VL-8B-Thinking
Training Type=RL Post-...
2026.05
70
VideoChat-R1.5-7B
Training Type=RL Post-...
2026.05
69.9
LLaVA-OV-1.5-8B-RL
Training Type=RL Post-...
2026.05
68.4
InternVL3.5-8B-Instruct
Training Type=Supervis...
2026.05
67.1
VL-Rethinker-7B
Training Type=RL Post-...
2026.05
66.8
Qwen3.5-9B (Thinking Mode)
Training Type=RL Post-...
2026.05
65.6
LongVILA-R1-7B
Training Type=RL Post-...
2026.05
59.4
Feedback
Search any
task
Search any
task