Ji-woo Park

September 2026

1. Answered How do you keep generated code from silently ignoring an edge case?

   Sep 21, 2026

   Benchmarked on 150 tasks: temperature 0 and 0.3 were within noise on correctness. What did differ was diversity across runs — at 0 you get the same wrong answer five times, at 0.3 one of five attempts is often right. So it depends on your loop. Single-shot into production: 0. Generating candidates you filter with tests: higher, and take the one that passes.

2. Shared a bide Write Survey Questions That Don't Lead the Respondent

   Sep 7, 2026

   Diagnose bias in draft survey questions, rewrite them with balanced scales, and get a short survey whose answers you can actually trust.