ssla_individual_defendants
- Task Description: This task requires the LLM to extract the identities of individual defendants from excerpts of securities class action complaints.
- Task Type: Extraction
- Document Type: complaint
- Number of Samples: 1015
- Input Length Range: 16-1710 tokens
- Evaluation Metrics: f1 (maximize)
- Tags: extraction, interpretation, securities 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 | f1 | Date | Results |
|---|---|---|---|---|
| 1 | meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo | 0.370 | 2025-07-25 | View |
| 2 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | 0.322 | 2025-08-04 | View |
| 3 | gpt-4o-mini | 0.295 | 2025-07-02 | View |
| 4 | gpt-4.1-nano | 0.197 | 2025-07-03 | View |
| 5 | claude-3-5-haiku-20241022 | 0.188 | 2025-08-01 | View |
| 6 | claude-3-haiku-20240307 | 0.002 | 2025-07-25 | View |
| 7 | google/gemma-2-27b-it | 0.001 | 2025-07-24 | View |