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

Selective Prediction on NCQA (test)

99.1PRR

Llama-3.1-8B

Updated 1d ago

Evaluation Results

Method	Links
Llama-3.1-8B 2026.06		99.1
Llama-3.1-8B 2026.06		99.1
Spectrum-Qwen3-14B 2026.06		69.5
A^2Search-7B 2026.06		68.3
A^2Search-7B 2026.06		67
Qwen3-14B 2026.06		64.5
A^2Search-7B 2026.06		63.8
A^2Search-7B 2026.06		63
Qwen3-14B 2026.06		60.6
Spectrum-Qwen3-14B 2026.06		59.9
Qwen3-14B 2026.06		57.7
Qwen3-14B 2026.06		53.4
Llama-3.1-8B 2026.06		50.7
Llama-3.1-8B 2026.06		50.2
Qwen3-14B 2026.06		49.9
Llama-3.1-8B 2026.06		44.2
Spectrum-Llama-3.1-8B 2026.06		43.3
Llama-3.1-8B 2026.06		36.9
Llama-3.1-8B 2026.06		33