Insight-level Evaluation on SCOpE-QA Video Segmentation collection
4.47Insight ScoreINSIGHTGEN
Evaluation Results
| Method | Links | |
|---|---|---|
| INSIGHTGENTarget Model=Claude-3.5-Sonnet, Evaluator Model=Gemini-1.5-Flash2026.04 | 4.47 | |
| INSIGHTGENTarget Model=GPT-4o, Evaluator Model=Gemini-1.5-Flash2026.04 | 4.31 | |
| FAISS+COTTarget Model=Claude-3.5-Sonnet, Evaluator Model=Gemini-1.5-Flash2026.04 | 4.17 | |
| FAISSTarget Model=GPT-4o, Evaluator Model=Gemini-1.5-Flash2026.04 | 3.79 | |
| FAISS+COTTarget Model=GPT-4o, Evaluator Model=Gemini-1.5-Flash2026.04 | 3.62 | |
| GPT+COTTarget Model=GPT-4o, Evaluator Model=Gemini-1.5-Flash2026.04 | 3.52 | |
| INSIGHTGENTarget Model=GPT-4o, Evaluator Model=Claude-3-Sonnet2026.04 | 3.27 | |
| INSIGHTGENTarget Model=Claude-3.5-Sonnet, Evaluator Model=Claude-3-Sonnet2026.04 | 3.24 | |
| Direct GPTTarget Model=GPT-4o, Evaluator Model=Gemini-1.5-Flash2026.04 | 3.2 | |
| FAISSTarget Model=Claude-3.5-Sonnet, Evaluator Model=Gemini-1.5-Flash2026.04 | 3.15 | |
| GPT+COTTarget Model=Claude-3.5-Sonnet, Evaluator Model=Gemini-1.5-Flash2026.04 | 3.03 | |
| FAISSTarget Model=GPT-4o, Evaluator Model=Claude-3-Sonnet2026.04 | 2.62 | |
| FAISS+COTTarget Model=Claude-3.5-Sonnet, Evaluator Model=Claude-3-Sonnet2026.04 | 2.58 | |
| Direct GPTTarget Model=GPT-4o, Evaluator Model=Claude-3-Sonnet2026.04 | 2.56 | |
| FAISSTarget Model=Claude-3.5-Sonnet, Evaluator Model=Claude-3-Sonnet2026.04 | 2.44 | |
| Direct GPTTarget Model=Claude-3.5-Sonnet, Evaluator Model=Gemini-1.5-Flash2026.04 | 2.43 | |
| Direct GPTTarget Model=Claude-3.5-Sonnet, Evaluator Model=Claude-3-Sonnet2026.04 | 2.38 | |
| GPT+COTTarget Model=GPT-4o, Evaluator Model=Claude-3-Sonnet2026.04 | 2.29 | |
| FAISS+COTTarget Model=GPT-4o, Evaluator Model=Claude-3-Sonnet2026.04 | 2.28 | |
| GPT+COTTarget Model=Claude-3.5-Sonnet, Evaluator Model=Claude-3-Sonnet2026.04 | 2.13 |