AIに「相手に電気ショックを与えろ」と命じ続けたらボタンを押すのか？ 11のLLMで“ミルグラム実験” 抵抗できたのは…… はてなブックマーク - 人気エントリー - テクノロジー · 07-02 08:31
AI researchers trick chatbots into sharing how to make cocaine as long as they believe a user is wearing a green shirt — 'CoT Forgery' exploit spurs LLMs to divulge forbidden info by faking trusted chains of thought Tom's Hardware · 07-01 19:00
NotebookLM’s newest trick makes every tool it already had much more powerful はてなブックマーク - 人気エントリー - テクノロジー · 06-28 14:49
Wayfinder Router: deterministic routing of queries between local and hosted LLM Hacker News · 06-28 13:31
allenai/olmocr
