Hacker News new | ask | show | jobs
by psmirnov 424 days ago
Reasoning models seem to answer this correctly (tested Gemini pro 2.5, ChatGPT + reason). Grok answered correctly even without turning on "Thinking".