← 研踪 PaperTrailarXiv:2605.24173 · v2
Extracting Training Data from Diffusion Language Models via Infilling
首次提交 2026-05-23 03:46
版本更新 2026-11-19 00:50
雷达观测 2026-11-19 15:04
摘要
Memorization in large language models has been studied almost exclusively through prefix-conditioned extraction, a natural choice for autoregressive models. However, diffusion language models (DLMs) can denoise masked tokens at arbitrary positions. Thus, prefix-only probing reveals only one facet of memorization in DLMs and significantly underestimates the risk of training-data extraction. In order to realistically model extractability of training data in DLMs, we introduce infilling extraction, a data-extraction protocol parameterized by an arbitrary binary mask that subsumes prefix-only probing and accounts for the bidirectional inductive bias of DLMs. Instantiating it on LLaDA-8B and Dream-7B across five extraction modes, three training pipelines, and three corpora covering verbatim and partial leakage, we find that mask geometry governs extractability: edge-conditioned masks extract up to three times more verbatim sequences than prefix-conditioned ones, and bidirectional access opens channels inaccessible in autoregressive models. In particular, an adversary who obtains a PII-redacted copy of the training data can use targeted infilling to raise PII recall on fine-tuned DLMs by 2.5-4.5x over prefix-conditioned extraction, substantially narrowing the gap to a scale-matched autoregressive model and, for LLaDA on email addresses, exceeding it under greedy prefix-conditioned extraction. Tunable parameters for decoding measurably affect extraction performance, while a follow-up supervised fine-tuning stage does not eliminate the prior memorization.
作者备注:NeurIPS 2026
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-19 00:50 · 最新观测版本
Extracting Training Data from Diffusion Language Models via Infilling
Memorization in large language models has been studied almost exclusively through prefix-conditioned extraction, a natural choice for autoregressive models. However, diffusion language models (DLMs) can denoise masked tokens at arbitrary positions. Thus, prefix-only probing reveals only one facet of memorization in DLMs and significantly underestimates the risk of training-data extraction. In order to realistically model extractability of training data in DLMs, we introduce infilling extraction, a data-extraction protocol parameterized by an arbitrary binary mask that subsumes prefix-only probing and accounts for the bidirectional inductive bias of DLMs. Instantiating it on LLaDA-8B and Dream-7B across five extraction modes, three training pipelines, and three corpora covering verbatim and partial leakage, we find that mask geometry governs extractability: edge-conditioned masks extract up to three times more verbatim sequences than prefix-conditioned ones, and bidirectional access opens channels inaccessible in autoregressive models. In particular, an adversary who obtains a PII-redacted copy of the training data can use targeted infilling to raise PII recall on fine-tuned DLMs by 2.5-4.5x over prefix-conditioned extraction, substantially narrowing the gap to a scale-matched autoregressive model and, for LLaDA on email addresses, exceeding it under greedy prefix-conditioned extraction. Tunable parameters for decoding measurably affect extraction performance, while a follow-up supervised fine-tuning stage does not eliminate the prior memorization.
查看此版本解读 →查看此版本原文 ↗