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

Question Answering on MedQA, PubMedQA, MedMCQA, and HeadQA Aggregate

50.63Average Accuracy

OLMo-2-7B-Instruct

Updated 10d ago

Evaluation Results

Method	Links
OLMo-2-7B-Instruct 2026.04		50.63
DPO 2026.04		50.09
DPO 2026.04		48.8
SKU 2026.04		48.17
RetainOnlyFT 2026.04		48
EU 2026.04		47.53
Unlearning 2026.04		46.95
Unlearning 2026.04		46.33
Eraser 2026.04		45.82
Eraser 2026.04		45
RetainOnlyFT 2026.04		44.96
EU 2026.04		44.68
Llama-3.2-3B-Instruct 2026.04		42.9
SKU 2026.04		41.98
Unlearning 2026.04		38.98
RetainOnlyFT 2026.04		38.2
Eraser 2026.04		37.64
Eraser 2026.04		37.2
SKU 2026.04		37.13
DPO 2026.04		37.12
Llama-3.2-1B-Instruct 2026.04		36.72
EU 2026.04		36.05
RetainOnlyFT 2026.04		35.67
Unlearning 2026.04		35.59
EU 2026.04		35.41
DPO 2026.04		35.07
OLMo-2-1B-Instruct 2026.04		33.86
SKU 2026.04		33.14