
Description
General benchmarks can't tell whether AI agents handle a lawyer's daily work. Harvey Labs is legal AI company Harvey's benchmark for evaluating and improving agents on legal work.
Its tasks mirror real legal practice, measuring research, drafting and analysis.
Legal tasks:Realistic.
Evaluation:Multi-dimensional.
Improvement:Find gaps.
Its tasks mirror real legal practice, measuring research, drafting and analysis.
Features
Legal tasks:Realistic.
Evaluation:Multi-dimensional.
Improvement:Find gaps.
