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

Mathematical Reasoning on AMC 24 (Pass@128)

79Pass@128

Qwen2.5-7B + GRPO w/ VERL.

Updated 5d ago

Evaluation Results

Method	Links
Qwen2.5-7B + GRPO w/ VERL. 2025.09		79
Qwen2.5-7B + GRPO w/ VERL. 2025.09		79
Qwen2.5-7B + GRPO 2025.09		78.3
Qwen2.5-7B + GRPO 2025.09		78.3
Mathstral-7B-v0.1 + GRPO w/ VERL. 2025.09		76.9
Qwen2.5-7B + PPO w/ VERL. 2025.09		74.4
Qwen2.5-7B + PPO w/ VERL. 2025.09		74.4
Qwen2.5-7B + PPO 2025.09		74.3
Qwen2.5-7B + PPO 2025.09		74.3
Qwen2.5-7B 2025.09		73.7
Qwen2.5-7B 2025.09		73.7
Mathstral-7B-v0.1 + PPO 2025.09		70.7
Mathstral-7B-v0.1 + PPO w/ VERL. 2025.09		69.9
Mathstral-7B-v0.1 + GRPO 2025.09		69.2
Qwen2.5-3B 2025.09		69
Qwen2.5-3B + GRPO 2025.09		68.5
Qwen2.5-3B + PPO 2025.09		67.9
Qwen2.5-3B + GRPO w/ VERL. 2025.09		67.8
Qwen2.5-3B + PPO w/ VERL. 2025.09		67.3
Llama-3.1-8B-Instruct + GRPO w/ VERL. 2025.09		63.1
Mathstral-7B-v0.1 2025.09		60.9
Llama-3.2-3B-Instruct + GRPO 2025.09		60.6
Llama-3.2-3B-Instruct + GRPO 2025.09		60.6
Llama-3.1-8B-Instruct + PPO w/ VERL. 2025.09		60
Llama-3.2-3B-Instruct + GRPO w/ VERL. 2025.09		59
Llama-3.2-3B-Instruct + GRPO w/ VERL. 2025.09		59
Llama-3.1-8B-Instruct + PPO 2025.09		59
Llama-3.2-3B-Instruct + PPO w/ VERL. 2025.09		57.8
Llama-3.2-3B-Instruct + PPO w/ VERL. 2025.09		57.8
Llama-3.1-8B-Instruct 2025.09		57.4
Llama-3.2-3B-Instruct + PPO 2025.09		57
Llama-3.2-3B-Instruct + PPO 2025.09		57
Llama-3.1-8B-Instruct + GRPO 2025.09		56.9
Llama-3.2-3B-Instruct 2025.09		51.5
Llama-3.2-3B-Instruct 2025.09		51.5
Mistral-7B-v0.3 2025.09		39.6
Mistral-7B-v0.3 + GRPO w/ VERL. 2025.09		38
Mistral-7B-v0.3 + GRPO 2025.09		36
Mistral-7B-v0.3 + PPO w/ VERL. 2025.09		30.1
Mistral-7B-v0.3 + PPO 2025.09		25.1