maud_specific_performance
- Task Description: This task requires the model to read an excerpt from a merger agreement and select the correct wording of the Specific Performance clause regarding the parties’ entitlement in the event of a contractual breach.
- Task Type: Binary classification
- Document Type: merger agreement
- Number of Samples: 179
- Input Length Range: 38-400 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 | google/gemma-2-27b-it | 0.983 | 0.991 | 0.944 | 0.983 | 1.000 | 2025-07-24 | View |
| 2 | claude-3-haiku-20240307 | 0.972 | 0.879 | 0.892 | 0.972 | 1.000 | 2025-07-25 | View |
| 3 | claude-3-5-haiku-20241022 | 0.955 | 0.976 | 0.870 | 0.955 | 1.000 | 2025-08-01 | View |
| 4 | gpt-4o-mini | 0.944 | 0.970 | 0.845 | 0.944 | 1.000 | 2025-07-02 | View |
| 5 | gpt-4.1-nano | 0.927 | 0.500 | 0.481 | 0.927 | 1.000 | 2025-07-03 | View |
| 6 | meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo | 0.904 | 0.948 | 0.775 | 0.904 | 1.000 | 2025-07-25 | View |
| 7 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | 0.596 | 0.746 | 0.487 | 0.596 | 1.000 | 2025-08-03 | View |