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

Multi-task Language Understanding on MMLU (Generic Score)

63.73MMLU Score

MONA

Updated 5d ago

Evaluation Results

Method	Links
MONA 2026.05		63.73
Muon 2026.05		62.81
AdamW 2026.05		62.18
Llama-8B (unpruned) 2026.05		0.667
Llama-8B (unpruned) 2026.05		0.667
Qwen-3B (unpruned) 2026.05		0.644
Qwen-3B (unpruned) 2026.05		0.644
Qwen-3B (KOFF) 2026.05		0.61
Llama-8B (KOFF) 2026.05		0.603
Llama-3B (unpruned) 2026.05		0.56
Llama-3B (unpruned) 2026.05		0.56
Llama-3B (KOFF) 2026.05		0.52
Llama-3B (KOFF) 2026.05		0.484
Llama-8B (KOFF) 2026.05		0.415
Qwen-3B (KOFF) 2026.05		0.413
Llama-8B (pruning only) 2026.05		0.408
Llama-8B (pruning only) 2026.05		0.384
Qwen-3B (pruning only) 2026.05		0.329
Qwen-3B (pruning only) 2026.05		0.326
Llama-3B (pruning only) 2026.05		0.316
Llama-3B (pruning only) 2026.05		0.249