Question Answering on IntraBench (test)
75.9AccuracyIntrAgent
Evaluation Results
| Method | Links | |
|---|---|---|
| IntrAgentBackbone Model=Gemini-2.5 Pro2026.04 | 75.9 | |
| IntrAgentBackbone Model=GPT-4.12026.04 | 75.8 | |
| IntrAgentBackbone Model=DS-R12026.04 | 74.4 | |
| IntrAgentBackbone Model=o4-mini2026.04 | 73.8 | |
| IntrAgentBackbone Model=o32026.04 | 73.4 | |
| IntrAgentBackbone Model=GPT-4o2026.04 | 70 | |
| IntrAgentBackbone Model=Llama-3.1-70B2026.04 | 68.8 | |
| LongRAGBackbone Model=DS-R12026.04 | 65.5 | |
| LongRAGBackbone Model=GPT-4.12026.04 | 64.7 | |
| Vanilla RAG all-MiniLM-L6-v2Backbone Model=DS-R12026.04 | 64.3 | |
| Vanilla RAG E5-mistral-7b-instructBackbone Model=GPT-4.12026.04 | 64.2 | |
| Vanilla RAG E5-mistral-7b-instructBackbone Model=DS-R12026.04 | 63.8 | |
| Context. RAG E5-mistral-7b-instructBackbone Model=GPT-4.12026.04 | 63.8 | |
| SciMasterBackbone Model=DS-R12026.04 | 63.3 | |
| Vanilla RAG GritLM-7BBackbone Model=GPT-4.12026.04 | 63.2 | |
| Vanilla RAG GritLM-7BBackbone Model=DS-R12026.04 | 63.2 | |
| Context. RAG E5-mistral-7b-instructBackbone Model=DS-R12026.04 | 62.8 | |
| Context. RAG GritLM-7BBackbone Model=GPT-4.12026.04 | 62.8 | |
| Agentic-Hybrid-RAGBackbone Model=DS-R12026.04 | 62.3 | |
| LongRAGBackbone Model=GPT-4o2026.04 | 62.1 | |
| Vanilla RAG all-MiniLM-L6-v2Backbone Model=Gemini-2.5 Pro2026.04 | 61.8 | |
| Context. RAG GritLM-7BBackbone Model=DS-R12026.04 | 61.6 | |
| Context. RAG GritLM-7BBackbone Model=Gemini-2.5 Pro2026.04 | 61.6 | |
| Vanilla RAG all-MiniLM-L6-v2Backbone Model=o4-mini2026.04 | 61.5 | |
| R2AGBackbone Model=DS-R12026.04 | 61.5 | |
| Vanilla RAG E5-mistral-7b-instructBackbone Model=o4-mini2026.04 | 61.4 | |
| Vanilla RAG GritLM-7BBackbone Model=Llama-3.1-70B2026.04 | 61.4 | |
| Vanilla RAG all-MiniLM-L6-v2Backbone Model=GPT-4.12026.04 | 61.2 | |
| Context. RAG GritLM-7BBackbone Model=GPT-4o2026.04 | 60.8 | |
| Context. RAG E5-mistral-7b-instructBackbone Model=GPT-4o2026.04 | 60.7 | |
| Context. RAG GritLM-7BBackbone Model=o4-mini2026.04 | 60.7 | |
| Vanilla RAG E5-mistral-7b-instructBackbone Model=Llama-3.1-70B2026.04 | 60.5 | |
| Vanilla RAG all-MiniLM-L6-v2Backbone Model=o32026.04 | 60.4 | |
| Vanilla RAG GritLM-7BBackbone Model=GPT-4o2026.04 | 60.4 | |
| Vanilla RAG all-MiniLM-L6-v2Backbone Model=GPT-4o2026.04 | 60.3 | |
| Vanilla RAG E5-mistral-7b-instructBackbone Model=o32026.04 | 60.3 | |
| Agentic-Hybrid-RAGBackbone Model=GPT-4.12026.04 | 60.2 | |
| Vanilla RAG E5-mistral-7b-instructBackbone Model=Gemini-2.5 Pro2026.04 | 59.9 | |
| Agentic-Hybrid-RAGBackbone Model=GPT-4o2026.04 | 59.8 | |
| Vanilla RAG GritLM-7BBackbone Model=o32026.04 | 59.7 | |
| R2AGBackbone Model=GPT-4.12026.04 | 59.5 | |
| Vanilla RAG E5-mistral-7b-instructBackbone Model=GPT-4o2026.04 | 59.4 | |
| R2AGBackbone Model=GPT-4o2026.04 | 59.4 | |
| Vanilla RAG all-MiniLM-L6-v2Backbone Model=Llama-3.1-70B2026.04 | 59.2 | |
| Context. RAG GritLM-7BBackbone Model=Llama-3.1-70B2026.04 | 59.2 | |
| Context. RAG E5-mistral-7b-instructBackbone Model=o32026.04 | 59.1 | |
| SciMasterBackbone Model=GPT-4o2026.04 | 59 | |
| Context. RAG E5-mistral-7b-instructBackbone Model=Gemini-2.5 Pro2026.04 | 58.9 | |
| Context. RAG E5-mistral-7b-instructBackbone Model=Llama-3.1-70B2026.04 | 58.9 | |
| Vanilla RAG GritLM-7BBackbone Model=o4-mini2026.04 | 58.4 | |
| Vanilla RAG GritLM-7BBackbone Model=Gemini-2.5 Pro2026.04 | 58.4 | |
| Context. RAG GritLM-7BBackbone Model=o32026.04 | 58.4 | |
| Context. RAG E5-mistral-7b-instructBackbone Model=o4-mini2026.04 | 58.3 | |
| LongRAGBackbone Model=o4-mini2026.04 | 58.3 | |
| SciMasterBackbone Model=o4-mini2026.04 | 58.1 | |
| Agentic-Hybrid-RAGBackbone Model=o4-mini2026.04 | 57.8 | |
| SciMasterBackbone Model=GPT-4.12026.04 | 57.6 | |
| Agentic-Hybrid-RAGBackbone Model=o32026.04 | 57.5 | |
| LongRAGBackbone Model=Llama-3.1-70B2026.04 | 57.4 | |
| Agentic-Hybrid-RAGBackbone Model=Gemini-2.5 Pro2026.04 | 57.2 | |
| SciMasterBackbone Model=o32026.04 | 57.2 | |
| SciMasterBackbone Model=Gemini-2.5 Pro2026.04 | 57.2 | |
| LongRAGBackbone Model=Gemini-2.5 Pro2026.04 | 57.1 | |
| LongRAGBackbone Model=o32026.04 | 57 | |
| SciMasterBackbone Model=Llama-3.1-70B2026.04 | 57 | |
| R2AGBackbone Model=o32026.04 | 56.6 | |
| Agentic-Hybrid-RAGBackbone Model=Llama-3.1-70B2026.04 | 56.6 | |
| LUMOSBackbone Model=o4-mini2026.04 | 56.4 | |
| R2AGBackbone Model=Llama-3.1-70B2026.04 | 56.1 | |
| R2AGBackbone Model=Gemini-2.5 Pro2026.04 | 55.6 | |
| LUMOSBackbone Model=DS-R12026.04 | 55.4 | |
| R2AGBackbone Model=o4-mini2026.04 | 55.3 | |
| LUMOSBackbone Model=o32026.04 | 55.2 | |
| LUMOSBackbone Model=Gemini-2.5 Pro2026.04 | 54.9 | |
| LUMOSBackbone Model=Llama-3.1-70B2026.04 | 54.4 | |
| PaperQA2Backbone Model=DS-R12026.04 | 54 | |
| PaperQA2Backbone Model=Llama-3.1-70B2026.04 | 53.8 | |
| LUMOSBackbone Model=GPT-4.12026.04 | 52.1 | |
| PaperQA2Backbone Model=o32026.04 | 51.8 | |
| PaperQA2Backbone Model=Gemini-2.5 Pro2026.04 | 51.2 | |
| LUMOSBackbone Model=GPT-4o2026.04 | 50.2 | |
| PaperQA2Backbone Model=o4-mini2026.04 | 49.2 | |
| PaperQA2Backbone Model=GPT-4.12026.04 | 48.9 | |
| PaperQA2Backbone Model=GPT-4o2026.04 | 47.7 | |
| DRAGINBackbone Model=DS-R12026.04 | 46.9 | |
| DRAGINBackbone Model=o4-mini2026.04 | 46.9 | |
| DRAGINBackbone Model=Gemini-2.5 Pro2026.04 | 45.9 | |
| DRAGINBackbone Model=Llama-3.1-70B2026.04 | 45.4 | |
| DRAGINBackbone Model=GPT-4.12026.04 | 44.6 | |
| DRAGINBackbone Model=o32026.04 | 44 | |
| DRAGINBackbone Model=GPT-4o2026.04 | 42.5 |