按年被提到
20242024 · 1 次2026
AI 安全、伦理与社会影响›模型窃取、记忆与水印·arXiv:2403.06634 · cs.CR

Stealing Part of a Production Language Model

Nicholas Carlini, Daniel Paleka, Krishnamurthy Dj Dvijotham, Thomas Steinke, Jonathan Hayase, A. Feder Cooper, Katherine Lee, Matthew Jagielski, Milad Nasr, Arthur Conmy, Itay Yona, Eric Wallace 等 14 人 · 2024
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 2 人首次 2024-03-12 · 最近 2025-10-05
2不同的人
2活跃月份
2讨论串
34被引(OpenAlex)
arXiv 摘要(候选层不写导读)

We introduce the first model-stealing attack that extracts precise, nontrivial information from black-box production language models like OpenAI's ChatGPT or Google's PaLM-2. Specifically, our attack recovers the embedding projection layer (up to symmetries) of a transformer model, given typical API access. For under \$20 USD, our attack extracts the entire projection matrix of OpenAI's Ada and Babbage language models. We thereby confirm, for the first time, that these black-box models have a hidden dimension of 1024 and 2048, respectively. We also recover the exact hidden dimension size of the gpt-3.5-turbo model, and estimate it would cost under $2,000 in queries to recover the entire projection matrix. We conclude with potential defenses and mitigations, and discuss the implications of possible future work that could extend our attack.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论tptacek · 2025-10

“Yes. https://arxiv.org/abs/2403.06634 https://arxiv.org/abs/2311.17035 (I just have these ones off the top of my head because I'm a Nicholas Carlini fan and we interviewed him about these attacks.)”

What GPT-OSS leaks about OpenAI's training data原评论 ↗
被提交为 story · 1 次