{"id":"346643ff-286d-4ae5-9f91-40dbf25c65fa","url":"https://api.tumn.com/news/346643ff-286d-4ae5-9f91-40dbf25c65fa","title":"AI Models Can Provably Hide Arbitrary Capabilities","summary":"Comments","sections":[],"faq":[],"sources":[{"url":"https://arxiv.org/abs/2610.13928","title":"Hacker News 프런트페이지","publishedAt":"2026-10-13T02:07:08.000Z"}],"author":"TUMN 큐레이션","aiAssisted":false,"reviewStatus":"automated","reviewNote":"원문 제목과 공개된 발췌를 인용하고 원문으로 연결합니다. 내용을 재작성하거나 요약하지 않았습니다.","publishedAt":"2026-10-13T02:07:08.000Z","modifiedAt":"2026-10-13T08:24:50.000Z","checkedAt":"2026-10-13T08:24:50.733Z","language":"ko","kind":"link","sourceLabel":"개발","profile":null,"via":null,"formats":{"json":"https://api.tumn.com/news/346643ff-286d-4ae5-9f91-40dbf25c65fa.json","markdown":"https://api.tumn.com/news/346643ff-286d-4ae5-9f91-40dbf25c65fa.md"}}