List your company

Mathematical Reasoning on AMC23 (Accuracy (%))

72.5Accuracy (%)MCTS (Vanilla)
46.553.256066.75Apr 1, 2026

Evaluation results

MethodLinks
MCTS (Vanilla)Base Model=Qwen2026.0472.5
MCTS + PEBase Model=Qwen2026.0469.2
Beam SearchBase Model=Qwen2026.0467.5
Adaptive Parallel Monte Carlo Tree Search (Full Suite)Base Model=Qwen2026.0465
MCTS (Vanilla)Base Model=Llama2026.0457.5
Adaptive Parallel Monte Carlo Tree Search (Full Suite)Base Model=Llama2026.0455
Beam SearchBase Model=Llama2026.0452.5
MCTS + PEBase Model=Llama2026.0447.5