List your company

Mathematical Problem Solving on Olympiad Bench

49.1Pass@8 AccuracyVI-CuRL
-0.40412.44825.338.152Feb 13, 2026

Evaluation results

MethodLinks
VI-CuRLModel=Qwen2.5-Math-7B, Verifier Setting=Majority Vote (w/o. Verifier)2026.0249.1
VI-CuRLModel=Qwen2.5-Math-7B, Verifier Setting=Entropy (w/o. Verifier)2026.0243.4
No CurriculumModel=Qwen2.5-Math-7B, Verifier Setting=Majority Vote (w/o. Verifier)2026.0241.9
No CurriculumModel=Qwen2.5-Math-7B, Verifier Setting=Entropy (w/o. Verifier)2026.0235.2
Base (No RL)Model=Qwen2.5-Math-7B, Verifier Setting=None2026.0233.9
No CurriculumModel=Llama3.2-3B-Instruct, Verifier Setting=Majority Vote (w/o. Verifier)2026.0224
VI-CuRLModel=Llama3.2-3B-Instruct, Verifier Setting=Majority Vote (w/o. Verifier)2026.0223.3
VI-CuRLModel=Llama3.2-3B-Instruct, Verifier Setting=Entropy (w/o. Verifier)2026.0221.8
Base (No RL)Model=Llama3.2-3B-Instruct, Verifier Setting=None2026.0216.4
No CurriculumModel=Llama3.2-3B-Instruct, Verifier Setting=Entropy (w/o. Verifier)2026.021.5