The paper addresses multi‑bit generative watermarking for large language models under a worst‑case false‑alarm constraint. It shows that a previously proposed scheme is suboptimal and introduces two new encoding‑decoding constructions that achieve the established lower bound on miss‑detection probability, fully characterizing optimal performance. The authors formulate the problem as a linear program, derive conditions for optimality, identify the failure mode of the earlier construction, and compare trade‑offs between the two new schemes.
arXiv:2508.03829v3 Announce Type: replace Abstract: The growing deployment of Large Language Models (LLMs) has raised concerns about their misuse in generating harmful or deceptive content. To addres...
TTMark introduces a pairwise watermarking framework that extends distortion‑free watermarking from single tokens to adjacent token pairs, enlarging the watermarking alphabet from V to V². By watermarking the joint distribution of consecutive tokens, the detector can exploit both token entropy and conditional entropy while maintaining distortion‑freeness. Experiments on multiple language models and datasets show that TTMARK improves detectability, robustness to edits, and localized watermark detection without degrading generation quality.
arXiv:2610.27653v1 Announce Type: cross Abstract: Watermarking has emerged as a key tool for identifying text generated by large language models (LLMs), with current schemes embedding statistical sig...
arXiv:2601. 11629v2 Announce Type: replace-cross Abstract: We demonstrate that while the current approaches for language model watermarking are effective for open-ended generation, they are inadequate at watermarking LM outputs for constrained generation tasks with low-entropy output spaces.
arXiv:2512. 13325v2 Announce Type: replace-cross Abstract: Securing digital text is becoming increasingly relevant due to the widespread use of large 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.
The paper introduces Dual-Embedding Watermarking (DEW), a semantic watermarking technique for large language models that combines contextual and token-level embeddings. DEW applies algebraic vector-space operations to generate a watermark signal that remains robust to paraphrasing and translation, while obfuscating the signal with pseudo-random matrices seeded by a secret key. Experiments demonstrate state‑of‑the‑art robustness, especially against translation, with minimal computational overhead and preserved text quality at lower watermark strengths.
WeaveMark is a new multi‑bit watermarking scheme for large language models that improves payload capacity, extraction accuracy, and text quality by using coded payload spreading, soft‑decision error‑correcting codes, and unbiased multilayer reweighting. It also adds zero‑bit layers for reliable detection of watermark presence. Experiments demonstrate significant gains, achieving an 89.8% match rate for 32‑bit messages at 200 tokens and maintaining 86.0% accuracy under 10% substitution attacks on 16‑bit messages, far outperforming the BiMark baseline.
The paper investigates latent‑space watermarking using pretrained generators, where a watermark encoder selects latent inputs based on a message and secret key to produce outputs with a specified conditional distribution. For finite alphabets, it derives inner and outer bounds on the rate–key trade‑off and characterizes the capacity region when the generator’s output uniquely determines the latent distribution. The study extends to jointly Gaussian models, identifies key sufficient statistics, optimally allocates secret‑key resources across modes, and analyzes robustness against regeneration attacks, providing compound capacity results and decay rates for repeated attacks.
The paper proposes a defense against watermark forgery in generative models by randomizing the selection of watermark keys for each query and accepting content only if exactly one key detects a watermark. This approach provides a sample‑count‑independent upper bound on forgery success for blind attackers while preserving model utility. Experiments show a dramatic drop in forgery success rates—from 87% to 1% with four keys for text and from 100% to 2% for images—without significant computational overhead.
arXiv:2608. 03174v1 Announce Type: cross Abstract: Generative AI systems increasingly produce content whose provenance is difficult to verify, motivating watermarking techniques for identifying model-generated outputs.
arXiv:2606. 18430v1 Announce Type: new Abstract: Statistical watermarks help organizations attribute large language model (LLM) outputs, yet existing detectors often struggle when watermark signals are weak, texts are repetitive, or watermarks are edited.