Causality on QRData
52AccuracyReAct + REFTOOL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ReAct + REFTOOLModel Backbone=GPT-4o2025.05 | 52 | — | |
| ReAct + REFTOOLModel Backbone=GPT-42025.05 | 51.3 | — | |
| ReActModel Backbone=GPT-42025.05 | 50.9 | — | |
| ReAct + RAGModel Backbone=GPT-4o2025.05 | 49.1 | — | |
| ReAct + REFTOOLModel Backbone=Gemini-1.5-Pro2025.05 | 48.3 | — | |
| ReAct + RAGModel Backbone=GPT-42025.05 | 48 | — | |
| ReActModel Backbone=Gemini-1.5-Pro2025.05 | 47.6 | — | |
| PoT + REFTOOLModel Backbone=GPT-4o2025.05 | 46.8 | — | |
| ReAct + RAGModel Backbone=Gemini-1.5-Pro2025.05 | 46.8 | — | |
| ReActModel Backbone=GPT-4o2025.05 | 46.5 | — | |
| PoT + REFTOOLModel Backbone=Gemini-1.5-Pro2025.05 | 43.9 | — | |
| PoT + RAGModel Backbone=GPT-4o2025.05 | 42 | — | |
| PoTModel Backbone=Gemini-1.5-Pro2025.05 | 41.3 | — | |
| CreatorModel Backbone=GPT-4o2025.05 | 39.8 | — | |
| PoTModel Backbone=GPT-4o2025.05 | 39.8 | — | |
| CreatorModel Backbone=GPT-42025.05 | 39.4 | — | |
| PoT + REFTOOLModel Backbone=GPT-42025.05 | 38.7 | — | |
| PoT + RAGModel Backbone=GPT-42025.05 | 37.5 | — | |
| PoT + REFTOOLModel Backbone=Llama-3.1-70B2025.05 | 36.8 | — | |
| TroVEModel Backbone=GPT-4o2025.05 | 36.4 | — | |
| PoT + RAGModel Backbone=Gemini-1.5-Pro2025.05 | 36.4 | — | |
| TroVEModel Backbone=GPT-42025.05 | 34.2 | — | |
| PoTModel Backbone=GPT-42025.05 | 34.2 | — | |
| LATMModel Backbone=Llama-3.1-70B2025.05 | 33.5 | — | |
| TroVEModel Backbone=Gemini-1.5-Pro2025.05 | 33.5 | — | |
| ReAct + REFTOOLModel Backbone=Llama-3.1-70B2025.05 | 33.5 | — | |
| PoTModel Backbone=Llama-3.1-70B2025.05 | 33.1 | — | |
| LATMModel Backbone=Gemini-1.5-Pro2025.05 | 32.3 | — | |
| ReAct + RAGModel Backbone=Llama-3.1-70B2025.05 | 32.3 | — | |
| ReActModel Backbone=Llama-3.1-70B2025.05 | 30.1 | — | |
| CreatorModel Backbone=Gemini-1.5-Pro2025.05 | 29.7 | — | |
| PoT + RAGModel Backbone=Llama-3.1-70B2025.05 | 29.7 | — | |
| LATMModel Backbone=GPT-42025.05 | 27 | — | |
| TroVEModel Backbone=Llama-3.1-70B2025.05 | 23.8 | — | |
| LATMModel Backbone=GPT-4o2025.05 | 20.9 | — | |
| CreatorModel Backbone=Llama-3.1-70B2025.05 | 14.9 | — | |
| Creator2025.05 | — | 31 | |
| LATM2025.05 | — | 28.4 | |
| PoT2025.05 | — | 37.1 | |
| PoT + RAG2025.05 | — | 36.4 | |
| PoT + REFTOOL2025.05 | — | 41.6 | |
| ReAct2025.05 | — | 43.8 | |
| ReAct + RAG2025.05 | — | 44.1 | |
| ReAct + REFTOOL2025.05 | — | 46.3 | |
| TroVE2025.05 | — | 32 |