Insight Generation on Internal non-scientific document collections Finance - Investment 2
4.73Set-level Score (Gemini 2.5 Flash)INSIGHTGEN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| INSIGHTGENBase Model=GPT-4o2026.04 | 4.73 | 3.5 | |
| INSIGHTGENBase Model=Claude-3.5-Sonnet2026.04 | 4.41 | 3.41 | |
| FAISS+CoTBase Model=Claude-3.5-Sonnet2026.04 | 4.09 | 2.68 | |
| GPT+CoTBase Model=Claude-3.5-Sonnet2026.04 | 3.82 | 2.73 | |
| FAISS+CoTBase Model=GPT-4o2026.04 | 3.46 | 2.23 | |
| GPT+CoTBase Model=GPT-4o2026.04 | 3.41 | 2.41 | |
| FAISSBase Model=GPT-4o2026.04 | 3.14 | 2.14 | |
| Direct GPTBase Model=GPT-4o2026.04 | 2.64 | 2.05 | |
| Direct GPTBase Model=Claude-3.5-Sonnet2026.04 | 2.46 | 1.68 | |
| FAISSBase Model=Claude-3.5-Sonnet2026.04 | 1.82 | 1.5 |