Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Chain-of-Thought Reasoning on UGPhysics AtomicPhysics
Loading...
15.1
Accuracy
MCNIG
7.716
9.633
11.55
13.467
Mar 18, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
MCNIG
Parameter scale=14B
2026.03
15.1
MCNIG
Parameter scale=8B
2026.03
14.7
IG
Parameter scale=8B
2026.03
13.7
QwenPRM
Parameter scale=7B
2026.03
13.4
Majority voting
2026.03
11.5
Granite PRM v2
2026.03
11
ImplicitPRM
2026.03
10.6
ORM
Parameter scale=8B
2026.03
10.4
MathShepherd
2026.03
10
OVM
2026.03
9.2
Single sampling
2026.03
8
Feedback
Search any
task
Search any
task