Y Hacker News new | ask | show | jobs

user: ag8 created: 2019-07-03 karma: 2006 runrl.com submissions: 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments Gourmand Syndrome 27 points | 9 comments 0 points | 0 comments guys why does armenian completely break Claude 99 points | 65 comments Sampling at negative temperature 203 points | 60 comments Perfectly Replicating Coca Cola [video] 1 points | 1 comments 0 points | 0 comments 0 points | 0 comments Po.ta.to 4 points | 2 comments Scaling pretraining affects RL sample efficiency 1 points | 0 comments Systematically generating tests that would have caught Anthropic's top‑K bug 2 points | 0 comments 0 points | 0 comments Tinker 4 points | 2 comments Training Qwen to answer briefly yet intelligently using feedback control 4 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments Launch HN: RunRL (YC X25) – Reinforcement learning as a service 71 points | 22 comments Generating the Funniest Joke with RL 1 points | 0 comments 0 points | 0 comments