arXiv AI September 24, 2026 By Bram Wouters

Optimizing watermarks for large language models

The paper "Optimizing watermarks for large language models" discusses the growing importance of watermarks for generative LLMs amid concerns about misuse. It presents a systematic multi‑objective optimization framework to balance watermark identifiability with the impact on text quality. The authors identify Pareto‑optimal solutions for a broad class of robust, efficient watermarks that outperform the current default watermark.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

llms

arXiv Machine Learning

Jul 8

Beyond Heuristic Tuning: Power-Calibrated LLM Watermarking

arXiv:2607. 05694v1 Announce Type: cross Abstract: Logit-based watermarking is a widely used mechanism for identifying LLM generated content, yet its effectiveness is governed by a fundamental trade-off between detectability and semantic distortion.

By Xiaopu Wang, Zelin He, Chengyuan Liu, Runze Li

llms

arXiv AI

Sep 18

Watermarking Diffusion Language Models

The paper introduces the first watermark designed specifically for diffusion language models (DLMs), which generate tokens in arbitrary order unlike traditional autoregressive models. It overcomes the challenge of missing prior tokens by applying the watermark in expectation over the context and promoting tokens that strengthen the watermark when used as context. Experiments show a >99% true positive rate with minimal quality loss and comparable robustness to existing autoregressive watermarks.

By Thibaud Gloaguen, Robin Staab, Nikola Jovanovi\'c, Martin Vechev

llms diffusion

arXiv AI

Jul 24

Making Open-Source Text LLM Watermarks Durable Against Merging

arXiv:2607. 20435v1 Announce Type: cross Abstract: Open-source LLMs (OSMs)arereaching near state-of-the-art performance, prompting prior works to trace the text they generate by embedding text watermarking algorithms directly into their weights.

By Luisa Scharff, Thibaud Gloaguen, Robin Staab, Martin Vechev

llms rag benchmarks safety

arXiv Computation and Language

Sep 14

Is Multilingual LLM Watermarking Truly Multilingual? Scaling Robustness to 100+ Languages via Back-Translation

arXiv:2510.18019v3 Announce Type: replace Abstract: Multilingual watermarking aims to make large language model (LLM) outputs traceable across languages, yet current methods still fall short. Despite...

By Asim Mohamed, Martin Gubri

llms nlp

arXiv Machine Learning

Sep 1

Optimal Estimation of Watermark Proportions in Hybrid AI-Human Texts

arXiv:2506.22343v2 Announce Type: replace-cross Abstract: Text watermarks in large language models (LLMs) are an increasingly important tool for detecting synthetic text and distinguishing human-writ...

By Xiang Li, Garrett Wen, Weiqing He, Jiayuan Wu, Qi Long, Weijie J. Su

llms

arXiv Machine Learning

Jul 9

Fast segmentation of watermarked texts from large language models through an epidemic change-point framework

arXiv:2509. 21160v2 Announce Type: replace-cross Abstract: With the growing use of large language models, concerns over content authenticity have spurred a variety of watermarking schemes.

By Soham Bonnerjee, Subhrajyoty Roy, Sayar Karmakar

llms computer-vision benchmarks

One email a morning, machine-written

Related stories

Beyond Heuristic Tuning: Power-Calibrated LLM Watermarking

Watermarking Diffusion Language Models

Making Open-Source Text LLM Watermarks Durable Against Merging

Is Multilingual LLM Watermarking Truly Multilingual? Scaling Robustness to 100+ Languages via Back-Translation

Optimal Estimation of Watermark Proportions in Hybrid AI-Human Texts

Fast segmentation of watermarked texts from large language models through an epidemic change-point framework