List your company

Question Answering on PIQA In-Domain (Accuracy)

89AccuracyFew-shot*
62.37669.28876.283.112Sep 26, 2025

Evaluation results

MethodLinks
Few-shot*Model=Qwen3-32B2025.0989
Few-shot*Model=Llama3.1-70B2025.0988.6
ICRModel=Llama3.1-70B2025.0983.2
ICRModel=Qwen3-32B2025.0982.8
Zero-shotModel=Qwen3-32B2025.0978
I2CLModel=Llama3.1-70B2025.0975.4
I2CLModel=Qwen3-32B2025.0974.2
FVModel=Llama3.1-70B2025.0971.2
Zero-shotModel=Llama3.1-70B2025.0970.4
FVModel=Qwen3-32B2025.0963.4