logoalt Hacker News

buppermintyesterday at 10:11 PM0 repliesview on HN

The paper title is a bit misleading. The tested detectors and models here are small and rather dated (Llama 3.1 8B and Gemini Flash 2.0 - these are basically in the level of a modern 1B model), and the actual paper says this only shows vulnerability in small model systems.