Share your thoughts, 1 month free Claude Pro on usSee more

Mathematical Reasoning on GSM-Hard (Acc, Pass@5)

48.57Accuracy (GSM-Hard)

ReLAR

Updated 1mo ago

Evaluation Results

Method	Links
ReLAR 2026.06		48.57	52.12
Qwen2.5-7B 2026.06		47.64	54.83
ReLAR 2026.06		45.02	47.58
Med42-Llama3-8B 2026.06		43.49	51.63
Llama-3-8B-Instruct 2026.06		43.28	51.84
Qwen2.5-Med-7B 2026.06		42.82	50.03
Qwen2.5-7B 2026.06		39.83	47.28
Med42-Mistral-7B 2026.06		38.84	47.23
Mistral-7B-v0.3 2026.06		38.64	46.28
Qwen2.5-Med-7B 2026.06		36.47	43.82
Mistral-7B-Instruct 2026.06		36.28	44.83
Med42-Llama3-8B 2026.06		35.62	42.93
MedGemma-4B 2026.06		34.63	42.08
Llama-3-8B-Instruct 2026.06		34.47	42.83
Gemma-7B 2026.06		31.84	39.47
Med42-Mistral-7B 2026.06		30.63	38.47
LLaMA-2-7B 2026.06		29.47	38.57
Mistral-7B-Instruct 2026.06		27.83	34.47
MedGemma-4B 2026.06		26.03	33.2
Mistral-7B-v0.3 2026.06		24.83	31.47
Gemma-7B 2026.06		22.47	29.63
LLaMA-2-7B 2026.06		22.34	27.83
Falcon-7B 2026.06		21.28	27.64
Falcon-7B 2026.06		14.47	19.83