Researchers develop a “divergence attack” that makes ChatGPT emit sequences copied from its training data, by prompting the LLM to repeat a word numerous times (Alex Ivanovs/Stack Diary)

https://www.techmeme.com/231129/p40

研究人員發展一種「偏離攻擊」,透過讓大型語言模型(ChatGPT)重複多次單詞,使其輸出從訓練資料中複製的序列。(Alex Ivanovs/Stack Diary)

via Techmeme

November 30, 2023 at 05:07AM

發佈留言

發佈留言必須填寫的電子郵件地址不會公開。 必填欄位標示為 *