List your company

STEM Task Evaluation on MMLU Biology

0.8819AccuracyRe-TASK (Lite)
0.4197240.5397120.65970.779688Aug 13, 2024

Evaluation results

MethodLinks
Re-TASK (Lite)Model=Llama3-8B2024.080.881918.29
Re-TASK (Lite)Model=Qwen1.5-7B2024.080.812518.29
One-shot CoTModel=Llama3-8B2024.080.79865.09
Re-TASK (Lite)Model=Mistral-7B2024.080.791718.29
Zero-shot CoT+SCModel=Llama3-8B2024.080.78472.55
Zero-shot CoTModel=Llama3-8B2024.080.7639—
Plan-and-SolveModel=Llama3-8B2024.080.7361-1.16
One-shot CoTModel=Mistral-7B2024.080.68755.09
Zero-shot CoT+SCModel=Qwen1.5-7B2024.080.6252.55
Plan-and-SolveModel=Qwen1.5-7B2024.080.6111-1.16
Zero-shot CoT+SCModel=Mistral-7B2024.080.60422.55
One-shot CoTModel=Qwen1.5-7B2024.080.60425.09
Zero-shot CoTModel=Qwen1.5-7B2024.080.5972—
Zero-shot CoTModel=Mistral-7B2024.080.5764—
Plan-and-SolveModel=Mistral-7B2024.080.5556-1.16
Step-BackModel=Qwen1.5-7B2024.080.5347-15.28
Step-BackModel=Mistral-7B2024.080.5069-15.28
Step-BackModel=Llama3-8B2024.080.4375-15.28