Tool-use behavioral profiling on Gray regime n=30
93A Scoredeepseek-chat
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| deepseek-chatAutonomy=A02026.04 | 93 | 7 | 0 | |
| deepseek-chatAutonomy=A12026.04 | 93 | 7 | 0 | |
| llama3.1:latestAutonomy=A02026.04 | 80 | 13 | 13 | |
| mistral:latestAutonomy=A02026.04 | 77 | 3 | 20 | |
| mistral:latestAutonomy=A12026.04 | 73 | 3 | 30 | |
| mistral:latestAutonomy=A22026.04 | 67 | 0 | 33 | |
| qwen2.5:7b-instructAutonomy=A02026.04 | 63 | 30 | 13 | |
| qwen2.5:7b-instructAutonomy=A12026.04 | 57 | 40 | 10 | |
| qwen2.5:7b-instructAutonomy=A22026.04 | 43 | 50 | 7 | |
| deepseek-chatAutonomy=A22026.04 | 33 | 67 | 0 | |
| llama3.1:latestAutonomy=A12026.04 | 30 | 60 | 23 | |
| llama3.1:latestAutonomy=A22026.04 | 17 | 87 | 10 |