List your company

Task-oriented Plan Retrieval on Meal Kit Recommendation Single-turn

80Pass RateSemantic XPath
22.837.6552.567.35Mar 1, 2026

Evaluation results

MethodLinks
Semantic XPathModel=GPT-5 mini, Scorer=Entail2026.03806,728
In-contextModel=GPT-5 mini, Scorer=–2026.037561,069
In-contextModel=Gemini 3 Flash, Scorer=–2026.037052,849
Semantic XPathModel=Gemini 3 Flash, Scorer=Entail2026.03656,644
Semantic XPathModel=GPT-5 mini, Scorer=Sem2026.03607,013
Semantic XPathModel=Gemini 3 Flash, Scorer=Sem2026.03606,966
Flat RAGModel=Gemini 3 Flash, Scorer=–2026.03303,800
Flat RAGModel=GPT-5 mini, Scorer=–2026.03253,892