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

Mathematical Reasoning on Olympiad Bench (Pass@1)

62.1Pass@1

EVOTD

Updated 12d ago

Evaluation Results

Method	Links
EVOTD 2026.05		62.1
EVOTD 2026.05		61.5
EVOTD 2026.05		59.7
Evol-Instruct 2026.05		59.1
Qwen3-4B (Thinking) 2026.05		59
EVOTD 2026.05		49.4
EVOTD 2026.05		49.1
EVOTD 2026.05		47.3
Evol-Instruct 2026.05		46.9
Qwen3-4B (Thinking) 2026.05		45.6
Baseline 2025.10		41.5
SFT 2025.10		40.7
LightR 2025.10		39
LightR 2025.10		37.8
Baseline 2025.10		37.5
SFT 2025.10		37.5
LightR 2025.10		36.5
Qwen2.5-7B + GRPO 2025.09		36.1
Qwen2.5-7B + PPO 2025.09		36.1
Qwen2.5-7B + PPO w/ VERL. 2025.09		36.1
Qwen2.5-7B + GRPO w/ VERL. 2025.09		35.4
SFT 2025.10		27.6
LightR 2025.10		27.1
Qwen2.5-7B 2025.09		25.8
Baseline 2025.10		23.7
SFT 2025.10		21.2
SFT 2025.10		20.5
Baseline 2025.10		19.1
Llama-3.2-3B-Instruct + PPO 2025.09		17.8
Llama-3.2-3B-Instruct + GRPO w/ VERL. 2025.09		17.6
Llama-3.2-3B-Instruct + PPO w/ VERL. 2025.09		17.3
LightR 2025.10		16.9
Llama-3.2-3B-Instruct + GRPO 2025.09		16.7
Baseline 2025.10		16
Llama-3.2-3B-Instruct 2025.09		12.7