List your company
ResearchFormerly WizWand

Mathematical Reasoning on HMMT February

0.4667Mean@64 AccSPS
0.2662920.3183210.370350.422379Feb 3, 2026Feb 15, 2026Feb 27, 2026Mar 12, 2026Mar 24, 2026Apr 5, 2026Apr 18, 2026
Updated 6mo ago

Evaluation results

MethodLinks
SPSBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.040.4667
GRPOBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.040.4
Agentic Proposingtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.365
PromptCot 2.0training budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.342
Polaristraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.338
Deepmathtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.337
DeepSeek-R1-Distill-Qwen-1.5BType=Base Model2026.040.3333
DAPOBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.040.3333
PromptCoTtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.331
MathSmithtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.329
OpenR1mathtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.318
Socratic-zerotraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.313
DeepSeek-V3.2-Spetraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.312
Qwen3-4B-Instruct-2507mode=zero-shot2026.020.31
Claude4.5-Opustraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.308
OpenthoughtsS3training budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.306
Qwen3-Maxtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.302
R-zerotraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.301
GSPOBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.040.3
GPT-5.2-Hightraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.299
NuminaMathtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.291
Gemini-3-Protraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.287
Wizardmathtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.278
Metamathtraining budget=10,000-trajectory, recipe=GRPO, backbone=Qwen3-4B-Instruct-25072026.020.274