maud_cor_standard_(intervening_event)
- Task Description: The model must select the answer that best characterizes the standard the board should follow when determining whether to change its recommendation in response to an intervening event in a merger agreement.
- Task Type: 9-way classification
- Document Type: merger agreement
- Number of Samples: 85
- Input Length Range: 80-1630 tokens
- Evaluation Metrics: accuracy (maximize), balanced_accuracy (maximize), f1_macro (maximize), f1_micro (maximize), valid_predictions_ratio (maximize)
- Tags: interpretation, merger agreement
- Paper: LegalBench: A Collaboratively Built Benchmark for Measuring Legal Reasoning in Large Language Models
- Dataset Download: https://hazyresearch.stanford.edu/legalbench/
7 submissions
| Rank | Model | accuracy | balanced_accuracy | f1_macro | f1_micro | valid_predictions_ratio | Date | Results |
|---|---|---|---|---|---|---|---|---|
| 1 | claude-3-5-haiku-20241022 | 0.667 | 0.572 | 0.511 | 0.667 | 1.000 | 2025-08-01 | View |
| 2 | gpt-4o-mini | 0.393 | 0.173 | 0.169 | 0.393 | 1.000 | 2025-07-02 | View |
| 3 | meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo | 0.286 | 0.189 | 0.232 | 0.286 | 1.000 | 2025-07-25 | View |
| 4 | gpt-4.1-nano | 0.226 | 0.100 | 0.118 | 0.226 | 1.000 | 2025-07-03 | View |
| 5 | google/gemma-2-27b-it | 0.202 | 0.091 | 0.098 | 0.202 | 1.000 | 2025-07-24 | View |
| 6 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | 0.143 | 0.099 | 0.136 | 0.143 | 1.000 | 2025-08-03 | View |
| 7 | claude-3-haiku-20240307 | 0.095 | 0.042 | 0.058 | 0.095 | 1.000 | 2025-07-28 | View |