Insight Generation on Internal document collections (non-scientific domain) Finance
4.67Insight Level ScoreINSIGHTGEN
Evaluation Results
| Method | Links | |
|---|---|---|
| INSIGHTGENTarget Model=GPT-4o, Evaluator=Gemini-2.5-Flash2026.04 | 4.67 | |
| INSIGHTGENTarget Model=Claude-3.5-Sonnet, Evaluator=Gemini-2.5-Flash2026.04 | 4.49 | |
| GPT+COTTarget Model=Claude-3.5-Sonnet, Evaluator=Gemini-2.5-Flash2026.04 | 4.39 | |
| FAISS+COTTarget Model=GPT-4o, Evaluator=Gemini-2.5-Flash2026.04 | 3.84 | |
| GPT+COTTarget Model=GPT-4o, Evaluator=Gemini-2.5-Flash2026.04 | 3.63 | |
| FAISSTarget Model=GPT-4o, Evaluator=Gemini-2.5-Flash2026.04 | 3.63 | |
| Direct GPTTarget Model=GPT-4o, Evaluator=Gemini-2.5-Flash2026.04 | 3.59 | |
| FAISS+COTTarget Model=Claude-3.5-Sonnet, Evaluator=Gemini-2.5-Flash2026.04 | 3.57 | |
| INSIGHTGENTarget Model=GPT-4o, Evaluator=Claude-4-Sonnet2026.04 | 3.37 | |
| INSIGHTGENTarget Model=Claude-3.5-Sonnet, Evaluator=Claude-4-Sonnet2026.04 | 3.24 | |
| GPT+COTTarget Model=Claude-3.5-Sonnet, Evaluator=Claude-4-Sonnet2026.04 | 3.06 | |
| Direct GPTTarget Model=Claude-3.5-Sonnet, Evaluator=Gemini-2.5-Flash2026.04 | 2.96 | |
| FAISS+COTTarget Model=Claude-3.5-Sonnet, Evaluator=Claude-4-Sonnet2026.04 | 2.67 | |
| GPT+COTTarget Model=GPT-4o, Evaluator=Claude-4-Sonnet2026.04 | 2.64 | |
| FAISS+COTTarget Model=GPT-4o, Evaluator=Claude-4-Sonnet2026.04 | 2.61 | |
| FAISSTarget Model=GPT-4o, Evaluator=Claude-4-Sonnet2026.04 | 2.47 | |
| Direct GPTTarget Model=GPT-4o, Evaluator=Claude-4-Sonnet2026.04 | 2.2 | |
| Direct GPTTarget Model=Claude-3.5-Sonnet, Evaluator=Claude-4-Sonnet2026.04 | 1.97 | |
| FAISSTarget Model=Claude-3.5-Sonnet, Evaluator=Gemini-2.5-Flash2026.04 | 1.63 | |
| FAISSTarget Model=Claude-3.5-Sonnet, Evaluator=Claude-4-Sonnet2026.04 | 1.07 |