Compliance evaluation on Compliance domain (test)
98AccuracyGroq-qwen3-32b
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Groq-qwen3-32bParameters=32B, Evaluation protocol=zero-shot prompting2026.03 | 98 | 79.7 | 3,008.6 | 0.003 | |
| Groq-kimi-k2Parameters=Unknown, Evaluation protocol=zero-shot prompting2026.03 | 98 | 80.5 | 764.2 | 0.002 | |
| LPF-SPNParameters=~50M, Evaluation protocol=task-specific training2026.03 | 97.8 | 1.4 | 14.8 | 0 | |
| Groq-llama-3.3-70bParameters=70B, Evaluation protocol=zero-shot prompting2026.03 | 95.9 | 81.6 | 1,578.7 | 0.004 | |
| Groq-gpt-oss-120bParameters=120B, Evaluation protocol=zero-shot prompting2026.03 | 93.9 | 81.3 | 1,541.7 | 0.006 |