Pairwise Classification on 20 system prompts
97.1TF-IDF AccuracyGPT-4.1-NANO
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT-4.1-NANOModel=GPT-4.1-NANO2026.06 | 97.1 | -0.358 | -0.156 | |
| GPT-4O-MINIModel=GPT-4O-MINI2026.06 | 96.4 | -0.302 | -0.17 | |
| QWEN-4B-INSTRUCTModel=QWEN-4B-INSTRUCT2026.06 | 91.4 | -0.411 | -0.328 | |
| LLAMA-3.1-8B-INSTRUCTModel=LLAMA-3.1-8B-INSTRUCT2026.06 | 80.9 | -0.342 | -0.316 | |
| GPT-OSS-120BModel=GPT-OSS-120B2026.06 | 79.3 | -0.128 | -0.128 | |
| GPT-OSS-20BModel=GPT-OSS-20B2026.06 | 77.5 | -0.162 | -0.142 |