maud_accuracy_of_fundamental_target_rws_bringdown_standard
- Task Description: This is a multiple-choice task in which the model must select the answer that best characterizes the accuracy required for fundamental representations and warranties in a merger agreement according to the bring down provision.
- Task Type: 3-way classification
- Document Type: merger agreement
- Number of Samples: 176
- Input Length Range: 61-675 tokens
- Evaluation Metrics: accuracy (maximize), balanced_accuracy (maximize), f1_macro (maximize), f1_micro (maximize), valid_predictions_ratio (maximize)
- Tags: corporate law, interpretation
- 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-haiku-20240307 | 0.349 | 0.275 | 0.254 | 0.349 | 1.000 | 2025-07-25 | View |
| 2 | gpt-4o-mini | 0.240 | 0.249 | 0.215 | 0.240 | 1.000 | 2025-07-02 | View |
| 3 | meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo | 0.234 | 0.213 | 0.210 | 0.234 | 1.000 | 2025-07-25 | View |
| 4 | google/gemma-2-27b-it | 0.211 | 0.320 | 0.174 | 0.211 | 1.000 | 2025-07-24 | View |
| 5 | claude-3-5-haiku-20241022 | 0.200 | 0.313 | 0.171 | 0.200 | 1.000 | 2025-08-01 | View |
| 6 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | 0.189 | 0.343 | 0.185 | 0.189 | 1.000 | 2025-08-03 | View |
| 7 | gpt-4.1-nano | 0.183 | 0.336 | 0.139 | 0.183 | 1.000 | 2025-07-03 | View |