Show HN: A new benchmark for testing LLMs for deterministic outputs from Hacker News on 2026-04-29 16:01 (#759PK) Comments