List your company

Mathematical Reasoning on MATH500 (Accuracy, Avg.)

75.5AccuracyCoT
7.3825.06542.7560.435May 27, 2026

Evaluation results

MethodLinks
CoTBackbone=Qwen2.5-7B-Instruct, Method Category=Non-RL Baselines2026.0575.567.83
Self-ConsistencyBackbone=Qwen2.5-7B-Instruct, Method Category=Non-RL Baselines2026.0567.662.42
TRACERBackbone=Qwen2.5-7B-Instruct, Method Category=Non-RL/RL Baseline Representative2026.0561.261.85
Sparse MADBackbone=Qwen2.5-7B-Instruct, Method Category=Non-RL Baselines2026.0547.855.8
Single-Agent GRPOBackbone=Qwen2.5-7B-Instruct, Method Category=RL Baselines2026.0547.846.49
MADBackbone=Qwen2.5-7B-Instruct, Method Category=Non-RL Baselines2026.0547.656.11
Single-Agent GSPOBackbone=Qwen2.5-7B-Instruct, Method Category=RL Baselines2026.0545.845.63
Single-Agent GRPOBackbone=Phi-3 Mini 4K Instruct, Method Category=RL Baselines2026.0543.646.63
MoABackbone=Qwen2.5-7B-Instruct, Method Category=Non-RL Baselines2026.0543.253.39
CoTBackbone=Phi-3 Mini 4K Instruct, Method Category=Non-RL Baselines2026.0541.352.2
Single-Agent GSPOBackbone=Phi-3 Mini 4K Instruct, Method Category=RL Baselines2026.054145.72
TRACERBackbone=Phi-3 Mini 4K Instruct, Method Category=Non-RL/RL Baseline Representative2026.054050.27
MADBackbone=Phi-3 Mini 4K Instruct, Method Category=Non-RL Baselines2026.0537.451.1
Sparse MADBackbone=Phi-3 Mini 4K Instruct, Method Category=Non-RL Baselines2026.0537.451.13
MoABackbone=Phi-3 Mini 4K Instruct, Method Category=Non-RL Baselines2026.0536.647.26
MAPoRLBackbone=Phi-3 Mini 4K Instruct, Method Category=RL Baselines2026.053648.93
Self-ConsistencyBackbone=Phi-3 Mini 4K Instruct, Method Category=Non-RL Baselines2026.0535.852.32
MAGRPOBackbone=Phi-3 Mini 4K Instruct, Method Category=RL Baselines2026.0534.238.52
MAGRPOBackbone=Qwen2.5-7B-Instruct, Method Category=RL Baselines2026.0532.734.73
MAPoRLBackbone=Qwen2.5-7B-Instruct, Method Category=RL Baselines2026.051038.19