Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Count on Count (test)
Loading...
99.8
Accuracy
Finetuning + KL
-3.992
22.954
49.9
76.846
May 15, 2025
Jul 14, 2025
Sep 13, 2025
Nov 13, 2025
Jan 13, 2026
Mar 15, 2026
May 15, 2026
Accuracy
Updated 16d ago
Evaluation Results
Method
Method
Links
Accuracy
Finetuning + KL
2025.05
99.8
Tracr Injection
2025.05
99.2
Ours (multi-object)
consistent-checkpoint=...
2026.05
72.97
VisionReasoner-7B
consistent-checkpoint=...
2026.05
69.5
Qwen2.5-VL-7B
re-implemented=true, c...
2026.05
67.9
Qwen2-VL-7B
re-implemented=true, c...
2026.05
48
No editing
2025.05
0
Feedback
Search any
task
Search any
task