Graph-based Question Answering on 23 Queries (Original)
63.2Mean F1LLM Planner
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LLM PlannerLLM Backbone=Claude Haiku 4.52026.06 | 63.2 | — | 5 | |
| Agentic RAGLLM Backbone=Claude Haiku 4.52026.06 | 57.7 | — | 3 | |
| Det. GraphRAGLLM Backbone=Claude Haiku 4.52026.06 | 57.4 | — | — | |
| LLM PlannerLLM Backbone=Claude Haiku 4.52026.06 | 55.7 | — | — | |
| Agentic RAGLLM Backbone=Claude Haiku 4.52026.06 | 48.4 | — | — | |
| Det. GraphRAGLLM Backbone=Claude Haiku 4.52026.06 | 47.2 | — | 1 | |
| TF-IDF RAG2026.06 | 34.9 | — | 0 | |
| TF-IDF RAG2026.06 | 30.6 | — | — |