Task-oriented Plan Retrieval on Travel Itinerary Single-turn
95Pass RateIn-context
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| In-contextModel=Gemini 3 Flash, Scorer=–2026.03 | 95 | 12,846 | |
| Semantic XPathModel=GPT-5 mini, Scorer=Entail2026.03 | 85 | 5,161 | |
| Semantic XPathModel=GPT-5 mini, Scorer=Sem2026.03 | 75 | 5,548 | |
| Semantic XPathModel=Gemini 3 Flash, Scorer=Entail2026.03 | 75 | 5,240 | |
| Semantic XPathModel=Gemini 3 Flash, Scorer=Sem2026.03 | 65 | 5,454 | |
| In-contextModel=GPT-5 mini, Scorer=–2026.03 | 55 | 12,136 | |
| Flat RAGModel=Gemini 3 Flash, Scorer=–2026.03 | 35 | 3,349 | |
| Flat RAGModel=GPT-5 mini, Scorer=–2026.03 | 25 | 3,137 |