Insight Generation on Internal non-scientific document collections Legal Business Analysis
4.65Set-level Score (Gemini-2.5-Flash Judge)INSIGHTGEN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| INSIGHTGENBase Model=GPT-4o2026.04 | 4.65 | 3.47 | |
| FAISS+CoTBase Model=Claude-3.5-Sonnet2026.04 | 4.65 | 2.82 | |
| INSIGHTGENBase Model=Claude-3.5-Sonnet2026.04 | 4.41 | 3.35 | |
| FAISS+CoTBase Model=GPT-4o2026.04 | 3.77 | 3 | |
| FAISSBase Model=GPT-4o2026.04 | 3.59 | 2.77 | |
| GPT+CoTBase Model=Claude-3.5-Sonnet2026.04 | 3.41 | 2.59 | |
| Direct GPTBase Model=GPT-4o2026.04 | 2.94 | 2 | |
| FAISSBase Model=Claude-3.5-Sonnet2026.04 | 2.47 | 2.29 | |
| Direct GPTBase Model=Claude-3.5-Sonnet2026.04 | 2 | 1.71 | |
| GPT+CoTBase Model=GPT-4o2026.04 | 1.12 | 0.77 |