view article Article The state-of-the-art in open-source AI for Swiss legal tasks joelniklaus • 21 days ago • 6
Running 2 When the Benchmark Is Just a Rubber Stamp ⚖ 2 How scaffolding alone scored 27% on legal review
Running 59 Don't Train the Model, Evolve the Harness 🌿 59 Evolving an agent's harness, not its model, on Harvey's LAB