Find the AI nobody approved.Learn more
Hacker News1d ago442 points214 commentslive
Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hard
A Memdoor workflow using five Qwen3.8-27B models achieved a mean score of 9.8/10 on LiveCodeBench Hard, outperforming the original harness (9.6) and Fable 5 (8.0) over five passes on ten problems.
Read the full story atgithub.comWhy this is in the Signal
LAXIMA AI Signal curates the highest-velocity stories across Hacker News, GitHub trending, and new Hugging Face / Replicate model releases — quality-filtered, deduplicated, and refreshed every four hours. This item surfaced from Hacker News with 442 points (by guregodevo). We link straight to the original source above — see the full live feed.
More AI Signal briefs
- HNOccupySaaS
- HNU.S. workers took home their smallest share of capital since at least 1947
- HNMathematics has changed. Let's act like it
- HNThe AI cost of JavaScript frameworks
- HNPatrick Stewart taught us what good men do with power
- HNLenovo's 'encrypted' fingerprint USB is just an unencrypted SD card in a case
- HNBrazilian Court Sanctions Lawyers for Prompt Injection in Court Filing
- HNPolice chief unplugs Flock cameras after company ignores contract cancellation