← 研踪 PaperTrailarXiv:2611.24109 · v1
PIVOT: Distribution-Preserving Multi-BitWatermarking with Minimax-Optimal Scoring
首次提交 2026-11-18 17:26
版本更新 2026-11-18 17:26
雷达观测 2026-11-19 15:04
摘要
As LLM-generated content becomes increasingly widespread, reliable attribution calls for multi-bit watermarks that embed recoverable messages without distorting the output distribution. We introduce PIVOT, a multi-bit watermark with model-free verification based on mirrored rank-interval sampling. At each generation step, PIVOT uses a keyed pseudorandom scalar and a secret permutation of the vocabulary. The message bit determines whether next-token probability intervals are arranged in forward or reverse secret-rank order. Sampling from these mirrored layouts preserves every token probability exactly while inducing an opposite-signed dependence between the scalar and the sampled token's secret rank. This dependence yields a pivotal rank statistic whose null distribution is invariant to the unknown next-token law, enabling model-free calibration and decoding. We formulate joint watermark detection and message recovery as a multiple-hypothesis testing problem, derive a moment-robust minimax-optimal detection score over a relaxed moment class, and establish finite-sample recovery guarantees. Extensive experiments on two representative large language models validate the effectiveness of PIVOT, demonstrating reliable watermark detection and message recovery while preserving the generator's output distribution.
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。
v1 · 2026-11-18 17:26 · 最新观测版本
PIVOT: Distribution-Preserving Multi-BitWatermarking with Minimax-Optimal Scoring
As LLM-generated content becomes increasingly widespread, reliable attribution calls for multi-bit watermarks that embed recoverable messages without distorting the output distribution. We introduce PIVOT, a multi-bit watermark with model-free verification based on mirrored rank-interval sampling. At each generation step, PIVOT uses a keyed pseudorandom scalar and a secret permutation of the vocabulary. The message bit determines whether next-token probability intervals are arranged in forward or reverse secret-rank order. Sampling from these mirrored layouts preserves every token probability exactly while inducing an opposite-signed dependence between the scalar and the sampled token's secret rank. This dependence yields a pivotal rank statistic whose null distribution is invariant to the unknown next-token law, enabling model-free calibration and decoding. We formulate joint watermark detection and message recovery as a multiple-hypothesis testing problem, derive a moment-robust minimax-optimal detection score over a relaxed moment class, and establish finite-sample recovery guarantees. Extensive experiments on two representative large language models validate the effectiveness of PIVOT, demonstrating reliable watermark detection and message recovery while preserving the generator's output distribution.
查看此版本解读 →查看此版本原文 ↗