人工智能2023 · 分享于 Hacker News
Scalable Extraction of Training Data from (Production) Language Models
探索《Scalable Extraction of Training Data from (Production) Language Models》的原文与 Hacker News 社区讨论，查看来源并加入阅读清单。
从论文，走向讨论
看看社区如何讨论这个想法。
来自 Hacker News 的原始讨论。不同的视角、疑问，以及值得继续思考的线索。
最多展示 100 条相关讨论。论文可能出现在评论中，并非原帖投稿内容。
在讨论中发现的引用
按 HN 条目顺序展示最近的 50 条引用来源。
Yink404 ·
Nightshade: An offensive tool for artists against AI art generators
590 赞同678 评论
Ytplrbv ·
Beyond self-attention: How a small language model predicts the next token
474 赞同85 评论
Yfi-le ·
What GPT-OSS leaks about OpenAI's training data
348 赞同82 评论
Yjonymo ·
The current state of the theory that GPL propagates to AI models
221 赞同309 评论
Ygeorgehill ·
Sorry, but a new prompt for GPT-4 is not a paper
15 赞同1 评论
Ywazokazi ·
Scalable Extraction of Training Data from (Production) Language Models
8 赞同0 评论
Yniyikiza ·
Extracting Training Data from LLMs
3 赞同1 评论
Ynabla9 ·
Scalable Extraction of Training Data from (Production) Language Models
1 赞同0 评论
Yrntn ·
Scalable Extraction of Training Data from (Production) Language Models
1 赞同0 评论