This is a heavily interactive web application, and JavaScript is required. Simple HTML interfaces are possible, but that is not what this is.
Post
Kent Beck
kentbeck.com
did:plc:2leyxott3wzm56eiysrlkv3p
Whenever I see 3 significant digits in a benchmark, especially of a qualitative task, I immediately get suspicious. "Sonnet 4 improved our SWE-bench agent single pass score from 60.6% to 70.6%" So, 60 to 70?
What kind of errors does it still make? What new errors does it make?
2025-05-23T12:34:59.050Z