Fluency Evaluation on SummEval
46.9AccuracyCollabEval
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| CollabEval2026.03 | 46.9 | 2.103 | 77.8 | 18 | 4.5 | 0 | 21.9 | 78.1 | |
| SonnetMode=Single-LLM2026.03 | 46.8 | 1 | 65.9 | 24 | 21.4 | 5 | 29.7 | 70.3 | |
| MistralMode=Single-LLM2026.03 | 45.8 | 1 | 86.7 | 13.3 | 0 | 0 | 25 | 75 | |
| HaikuMode=Single-LLM2026.03 | 13.8 | 1 | 75.9 | 22.3 | 6.2 | 0 | 30.2 | 69.8 |