Risk Classification on APPSI-139
96Micro F1Llama3-8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Llama3-8BPrompting Strategy=zero-shot2026.04 | 96 | 48.98 | |
| TCSI-pp-V2Mode=specialized fine-tuning2026.04 | 95.6 | 58.77 | |
| GPT-4o-miniPrompting Strategy=zero-shot2026.04 | 89.34 | 51.24 | |
| Qwen3-8BPrompting Strategy=zero-shot2026.04 | 85.53 | 49.3 | |
| Gemini-2.5Prompting Strategy=zero-shot2026.04 | 85.05 | 45.95 |