List your company

Mathematical Reasoning on Minerva (Acc@16)

26Acc@16Verify-Init
11.85615.52819.222.872Feb 7, 2026

Evaluation results

MethodLinks
Verify-InitBackbone=Qwen2.5-7B-Instruct2026.0226
Verify-AlterBackbone=Qwen2.5-7B-Instruct2026.0226
GenerateBackbone=Qwen2.5-7B-Instruct2026.0225.3
Mixed-TrainBackbone=Qwen2.5-7B-Instruct2026.0224.6
Verify-AlterBackbone=Qwen2.5-3B-Instruct2026.0218.7
Verify-InitBackbone=Qwen2.5-3B-Instruct2026.0218.3
Mixed-TrainBackbone=Qwen2.5-3B-Instruct2026.0217.6
GenerateBackbone=Qwen2.5-3B-Instruct2026.0217
Verify-AlterBackbone=Qwen2.5-1.5B-Instruct2026.0213.9
Verify-InitBackbone=Qwen2.5-1.5B-Instruct2026.0213
Mixed-TrainBackbone=Qwen2.5-1.5B-Instruct2026.0212.7
GenerateBackbone=Qwen2.5-1.5B-Instruct2026.0212.4