List your company

Task-oriented Interaction on Travel Itinerary Multi-turn

90Pass RateIn-context
17.236.15573.9Mar 1, 2026

Evaluation results

MethodLinks
In-contextBackbone=GPT-5 mini2026.039028,342
Semantic XPathBackbone=GPT-5 mini, Scorer=Entail2026.03805,820
In-contextBackbone=Gemini 3 Flash2026.036031,639
Semantic XPathBackbone=Gemini 3 Flash, Scorer=Sem2026.03605,451
Semantic XPathBackbone=Gemini 3 Flash, Scorer=Entail2026.03605,671
Semantic XPathBackbone=GPT-5 mini, Scorer=Sem2026.03505,664
Flat RAGBackbone=GPT-5 mini2026.03203,250
Flat RAGBackbone=Gemini 3 Flash2026.03203,164