Insight Generation on Internal non-scientific document collections (Responsible AI Consulting)
4.5Set-level Score (Gemini-2.5-Flash Judge)INSIGHTGEN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| INSIGHTGENBase Model=GPT-4o2026.04 | 4.5 | 3.31 | |
| INSIGHTGENBase Model=Claude-3.5-Sonnet2026.04 | 4.31 | 3.5 | |
| FAISSBase Model=GPT-4o2026.04 | 3.94 | 3.13 | |
| FAISS+CoTBase Model=Claude-3.5-Sonnet2026.04 | 3.88 | 2.5 | |
| Direct GPTBase Model=GPT-4o2026.04 | 3.81 | 2.56 | |
| GPT+CoTBase Model=Claude-3.5-Sonnet2026.04 | 3.25 | 2.13 | |
| GPT+CoTBase Model=GPT-4o2026.04 | 3 | 2.5 | |
| FAISS+CoTBase Model=GPT-4o2026.04 | 2.69 | 2.19 | |
| Direct GPTBase Model=Claude-3.5-Sonnet2026.04 | 2.63 | 1.44 | |
| FAISSBase Model=Claude-3.5-Sonnet2026.04 | 2.31 | 1.94 |