Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multiple-choice Question Answering on PubMedQA (Accuracy)
Loading...
63.62
Accuracy
DS2-INSTRUCT
12.972
26.121
39.27
52.419
Mar 13, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
DS2-INSTRUCT
Model Family=Qwen2.5
2026.03
63.62
InstructMix
Model Family=Qwen2.5
2026.03
62.15
Self-Instruct
Model Family=Qwen2.5
2026.03
61.47
ExploreInstruct
Model Family=Qwen2.5
2026.03
60.89
Zero-Shot
Model Family=Qwen2.5
2026.03
60.34
DS2-INSTRUCT
Model Family=Llama3
2026.03
48.56
DS2-INSTRUCT
Model Family=Mistral
2026.03
39.11
InstructMix
Model Family=Mistral
2026.03
31.48
ExploreInstruct
Model Family=Mistral
2026.03
28.35
InstructMix
Model Family=Llama3
2026.03
27.41
Self-Instruct
Model Family=Mistral
2026.03
26.83
Self-Instruct
Model Family=Llama3
2026.03
25.34
Zero-Shot
Model Family=Llama3
2026.03
24.92
ExploreInstruct
Model Family=Llama3
2026.03
24.77
Zero-Shot
Model Family=Mistral
2026.03
14.92
Feedback
Search any
task
Search any
task