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

Multi-task Language Understanding on MMLU-IT

81.5Accuracy

Qwen3-30B-A3B

Updated 1mo ago

Evaluation Results

Method	Links
Qwen3-30B-A3B 2026.03		81.5
Qwen3-30B-A3B 2026.03		81.5
Gpt-oss-20b-high 2026.03		80.4
Gpt-oss-20b-high 2026.03		80.2
Qwen3-14B 2026.03		77.6
Qwen3-14B 2026.03		77.6
gemma-3-12b-it 2026.03		69
gemma-3-12b-it 2026.03		68.5
gemma-2-9b-it 2026.03		67.2
gemma-2-9b-it 2026.03		66.5
EngGPT2-16B-A3B 2026.03		65.5
EngGPT2-16B-A3B 2026.03		65.5
Llama-3.1-8B-Instruct 2026.03		60.6
Llama-3.1-8B-Instruct 2026.03		60.6
Moonlight-16B-A3B-Instruct 2026.03		49.2
Moonlight-16B-A3B-Instruct 2026.03		49.2