List your company

Stir on scientific experiment dataset

74.4Success RateCAPER
28.43240.36652.364.234Feb 10, 2026

Evaluation results

MethodLinks
CAPER2026.0274.4
MP+MLM+RLBackbone=GPT-4o2026.0270.8
MLM+RLBackbone=GPT-4o2026.0270.5
MLM+RLBackbone=LLaMA3.2, Reasoning Strategy=w/o CoT2026.0239.6
LLM+RLBackbone=GPT-4o2026.0234.8
LLM+RLBackbone=Llama-3.1-8B, Reasoning Strategy=w/o CoT2026.0230.2