Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Open-ended Visual Question Answering on Med-CXR
Loading...
31.73
Accuracy
T2R-R3
15.3084
19.5717
23.835
28.0983
Jun 24, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
T2R-R3
Scale=2B
2026.06
31.73
Judge Reward
Scale=2B
2026.06
28
T2R-R2
Scale=2B
2026.06
27.51
Frozen Teacher
Scale=4B
2026.06
24.41
T2R-R3
Scale=4B
2026.06
23.23
T2R-R2
Scale=4B
2026.06
22.87
T2R-R1
Scale=2B
2026.06
22.5
Judge Reward
Scale=4B
2026.06
22.38
Frozen Teacher
Scale=2B
2026.06
21.74
Base Model
Scale=4B
2026.06
21.74
RLVR Only
Scale=4B
2026.06
21.34
T2R-R1
Scale=4B
2026.06
21.27
RLVR Only
Scale=2B
2026.06
16.87
Base Model
Scale=2B
2026.06
15.94
Feedback
Search any
task
Search any
task