successor_liability
- Task Description: Given a fact pattern, identify the type of successor liability present.
- Task Type: 4-way classification
- Document Type: fact pattern
- Number of Samples: 50
- Input Length Range: 51-177 tokens
- Evaluation Metrics: f1 (maximize)
- Tags: corporate law, rule application
- Paper: LegalBench: A Collaboratively Built Benchmark for Measuring Legal Reasoning in Large Language Models
- Dataset Download: https://hazyresearch.stanford.edu/legalbench/
19 submissions
| Rank | Model | f1 | Date | Results |
|---|---|---|---|---|
| 1 | grok-4-0709 | 0.558 | 2025-07-30 | View |
| 2 | gpt-5-2025-08-07 | 0.534 | 2025-08-08 | View |
| 3 | o3 | 0.503 | 2025-07-08 | View |
| 4 | claude-3-5-haiku-20241022 | 0.500 | 2025-08-01 | View |
| 5 | claude-opus-4-20250514 | 0.406 | 2025-07-25 | View |
| 6 | claude-opus-4-1-20250805 | 0.382 | 2025-08-05 | View |
| 7 | claude-3-haiku-20240307 | 0.381 | 2025-07-25 | View |
| 8 | claude-sonnet-4-20250514 | 0.352 | 2025-07-25 | View |
| 9 | google/gemma-2-27b-it | 0.349 | 2025-07-24 | View |
| 10 | o4-mini | 0.343 | 2025-07-08 | View |
| 11 | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo | 0.338 | 2025-08-04 | View |
| 12 | openai/gpt-oss-120b | 0.322 | 2025-08-05 | View |
| 13 | meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo | 0.315 | 2025-07-25 | View |
| 14 | deepseek-ai/DeepSeek-V3 | 0.271 | 2025-07-10 | View |
| 15 | gpt-4o-mini | 0.235 | 2025-07-02 | View |
| 16 | gpt-4.1-nano | 0.178 | 2025-07-03 | View |
| 17 | o3-mini | 0.140 | 2025-07-09 | View |
| 18 | grok-3-mini | 0.086 | 2025-07-29 | View |
| 19 | deepseek-ai/DeepSeek-R1 | 0.043 | 2025-07-11 | View |