canada_tax_court_outcomes
- Task Description: Classify whether an excerpt from a Canada Tax Court decision includes the outcome of the appeal, and if so, specify whether the appeal was allowed or dismissed.
- Task Type: 3-way classification
- Document Type: judicial text
- Number of Samples: 250
- Input Length Range: 64-1018 tokens
- Evaluation Metrics: accuracy (maximize), balanced_accuracy (maximize), f1_macro (maximize), f1_micro (maximize), valid_predictions_ratio (maximize)
- Tags: rhetorical understanding, tax law
- 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 | 0.984 | 0.964 | 0.975 | 0.984 | 1.000 | 2025-08-01 | View |
| 2 | gpt-4o-mini | 0.980 | 0.936 | 0.946 | 0.980 | 1.000 | 2025-07-02 | View |
| 3 | gpt-4.1-nano | 0.980 | 0.961 | 0.960 | 0.980 | 1.000 | 2025-07-03 | View |
| 4 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | 0.980 | 0.936 | 0.958 | 0.980 | 1.000 | 2025-07-28 | View |
| 5 | meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo | 0.971 | 0.870 | 0.912 | 0.971 | 0.996 | 2025-07-24 | View |
| 6 | google/gemma-2-27b-it | 0.971 | 0.880 | 0.896 | 0.971 | 0.996 | 2025-07-24 | View |
| 7 | claude-3-haiku-20240307 | 0.930 | 0.653 | 0.634 | 0.930 | 1.000 | 2025-07-25 | View |