Sensitive Information Classification on TorchSight-Eval-500-External Enron
87.5AccuracyClaude S4
Evaluation Results
| Method | Links | |
|---|---|---|
| Claude S42026.05 | 87.5 | |
| BeamQuantization=q4_K_M2026.05 | 83.8 | |
| Qwen base2026.05 | 81.2 | |
| Gemini 2.52026.05 | 77.5 | |
| GPT-52026.05 | 47.5 |
| Method | Links | |
|---|---|---|
| Claude S42026.05 | 87.5 | |
| BeamQuantization=q4_K_M2026.05 | 83.8 | |
| Qwen base2026.05 | 81.2 | |
| Gemini 2.52026.05 | 77.5 | |
| GPT-52026.05 | 47.5 |