Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Abstract Visual Reasoning on VCog-Bench
Loading...
42.4
CVR Score
Ours (RL)
22.536
27.693
32.85
38.007
Jan 31, 2026
CVR Score
Raven Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
CVR Score
Raven Score
Ours (RL)
Training=RL
2026.01
42.4
12.1
Ours (SFT)
Training=SFT
2026.01
40.8
11.3
Qwen2.5-VL-7B-Instruct
Parameters=7B, Variant...
2026.01
39.8
12.1
InternVL2.5-8B
Parameters=8B
2026.01
38.2
12.3
LLaVA-OneVision-Qwen2-7B
Parameters=7B
2026.01
36.7
13
Janus-Pro-7B
Parameters=7B
2026.01
34.5
12.5
MM-EurekaQwen-7B
Parameters=7B
2026.01
33.3
10.4
Thyme
2026.01
32.3
10.8
R1-onevision-RL
Training=RL
2026.01
23.3
23.4
Feedback
Search any
task
Search any
task