alyssamaru
Re:For Computer Use, the harness matters as much as the model
on Thursday September 24, 2026 @01:29PM (#49834012)
Yes, would love to hear the methodology!
Re:We made Playwright 2x faster and 80% more token efficient
on Friday September 18, 2026 @12:52PM (#49757020)
How much does the harness really matter for evals?