25,000 entries generated by WithinUsAI on 2026-01-04. The dataset includes 8,000 verification tasks (TRUE/FALSE + correction), 9,000 self-contained quantitative/structured reasoning tasks, and 8,000 definitions or micro-refreshers. It is hosted on HuggingFace.
Use Cases
- Training models for factual verification in legal documents based on TRUE/FALSE + correction tasks.
- Developing models for structured quantitative reasoning based on self-contained reasoning tasks.
- Building models for legal concept understanding based on definitions and micro-refreshers.
- Benchmarking AI performance on complex legal and policy text analysis.
Strengths
- Contains 25,000 total entries.
- Includes 8,000 verification tasks with corrections.
- Includes 9,000 structured reasoning tasks.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Last updated 2026-01-04 10:24:55; freshness should be verified.
Provenance
- Source
- WithinUsAI
- Freshness
- 2026-01-04 10:24:55