XINLI1997 commited on
Commit
dadc277
·
verified ·
1 Parent(s): 853ebb2

Model card: training-time check at the final checkpoint (step 240, 25.62%), per camera-ready Appendix R

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -38,7 +38,7 @@ GRPO-trained reference checkpoint from **WirelessMathBench-XL: An Auditable Benc
38
 
39
  Locked protocol: 2k-token answer budget, T = 0.6, hierarchical verifier (deterministic match, canonicalisation, GPT-4.1-mini fallback judge).
40
 
41
- Training-time check under greedy model selection (paper Tab. 7): base 12.37% → GRPO 25.12%. These values use a different protocol from the locked scores.
42
 
43
  ## Usage
44
 
 
38
 
39
  Locked protocol: 2k-token answer budget, T = 0.6, hierarchical verifier (deterministic match, canonicalisation, GPT-4.1-mini fallback judge).
40
 
41
+ Training-time check at the final checkpoint (step 240; paper Appendix R): base 12.37% → GRPO 25.62%. These values use a different protocol from the locked scores.
42
 
43
  ## Usage
44