Bug Hunt Bench: GPT-6 Astra has unique blind spots, prompting a new model-routing strategy

PawelHuryn · x · 2026-09-05

PawelHuryn's Bug Hunt Bench — blind-graded runs of frontier coding models against 105 real planted bugs in repos — shares new data and FAQ:

Related event: GPT-6 Astra Tops 105-Real-Bug Benchmark with 48 Fixes(10 posts)→

Original post →

More from coding & agent

coding & agent channel →