Recent work shows that standard greedy-decoding extraction methods for quantifying memorization in LLMs miss how extraction risk varies across sequences. Probabilistic extraction -- computing the probability of generating a target suffix given a prefix under a decoding scheme -- addresses this, but is tractable only for verbatim memorization, missing near-verbatim instances that pose similar privacy and copyright risks. Quantifying near-verbatim extraction risk is expensive: the set of near-verbatim suffixes is combinatorially large, and reliable Monte Carlo (MC) estimation can require ~100,000 samples per sequence. To mitigate this cost, we introduce decoding-constrained beam search, which yields deterministic lower bounds on near-verbatim extraction risk at a cost comparable to ~20 MC samples per sequence. Across experiments, our approach surfaces information invisible to verbatim methods: many more extractable sequences, substantially larger per-sequence extraction mass, and patterns in how near-verbatim extraction risk manifests across model sizes and types of text.
HalluTracer: Pre-Decoding Truthfulness Prediction via Depth-Averaged Probe-Logit
University of Technology Sydney(悉尼科技大学);City University of Macau(澳门城市大学);Meta;actAVA AI(actAVA人工智能公司);Microsoft AI(微软人工智能);Squirrel Ai Learning(松鼠人工智能学习公司);East China Normal University(华东师范大学)