Legal Statute Identification on ILSI random subset of 1000 cases (test)
46.8Micro-avg PrecisionAoS model
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| AoS modelEvaluation subset=Random 1000 cases2025.12 | 46.8 | 47.2 | 47 | 37.5 | 32.8 | 33.1 | 31.9 | |
| LLMPromptLLM Model=GPT-4o mini, Prompting Strategy=Zero-shot2025.12 | 44.8 | 36.6 | 40.3 | 34.7 | 24.6 | 26.3 | 24.7 | |
| LLMPromptLLM Model=GPT-4o mini, Prompting Strategy=CoT2025.12 | 44 | 37.5 | 40.5 | 34.7 | 25.3 | 26.9 | 25 | |
| LLMPromptLLM Model=Mistral-7B-Instruct-v0.3, Prompting Strategy=Zero-shot2025.12 | 43.6 | 25.3 | 32.1 | 32 | 16.3 | 19.5 | 18.5 | |
| LLMPromptLLM Model=Mistral-7B-Instruct-v0.3, Prompting Strategy=CoT2025.12 | 42.7 | 30.9 | 35.8 | 34.7 | 20.9 | 23.7 | 21.7 |