Segment Intrusion on Segment Intrusion Laptop domain 1.0 (test)
100SI-EHuman Performance
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Human Performance2026.05 | 100 | 97 | 99 | 87 | 95.75 | |
| o4-mini2026.05 | 99 | 88 | 97.5 | 82.18 | 91.67 | |
| gemini-2.5-flash-preview-04-172026.05 | 98.5 | 89.5 | 81.79 | 66.48 | 84.07 | |
| o3-mini2026.05 | 98 | 90.5 | 94.5 | 80.99 | 91 | |
| gpt-4o2026.05 | 97 | 85.5 | 95 | 77.34 | 88.71 | |
| qwen2.5-72b-instruct-turbo2026.05 | 96.76 | 83 | 96 | 69.14 | 86.22 | |
| claude-3.7-sonnet-202502192026.05 | 96.5 | 82.5 | 94.5 | 80.89 | 88.6 | |
| llama-3.3-70b-instruct-turbo2026.05 | 96 | 81 | 94 | 71.11 | 85.53 | |
| deepseek-v32026.05 | 95.5 | 83.5 | 91 | 71.74 | 85.44 | |
| gemini-2.0-flash2026.05 | 93.5 | 78 | 92.73 | 68.28 | 83.13 | |
| claude-3.5-haiku-202410222026.05 | 91.67 | 79.22 | 88.78 | 64.69 | 81.09 | |
| gemini-2.0-flash-lite2026.05 | 86.64 | 70.87 | 90.55 | 67.93 | 79 | |
| qwen2.5-7b-instruct-turbo2026.05 | 85.15 | 63.5 | 60 | 30.14 | 59.7 | |
| llama-4-maverick-17b-128e-instruct-fp8Precision=fp82026.05 | 54.59 | 44.5 | 30.98 | 21.29 | 37.84 | |
| llama-3.2-3b-instruct-turbo2026.05 | 26.25 | 25.1 | 9.21 | 7.87 | 17.11 |