List your company

Commonsense Multi-hop QA on StrategyQA

8.1Full Length CoverageMinimal-core extraction
7.3727.5617.757.939May 14, 2026

Evaluation results

MethodLinks
Minimal-core extractionModel=GPT-52026.058.13.240608095
Minimal-core extractionModel=DeepSeek-R1-Distill-Qwen-32B2026.057.83.444567693
Minimal-core extractionModel=Qwen3-32B2026.057.53.547537391
Minimal-core extractionModel=DeepSeek-R1-Distill-Qwen-7B2026.057.43.750506988