14 papers
AutoMark: Enabling Autoresearch to Discover Better LLM Watermarks
Thibaud Gloaguen, Robin Staab, Martin Vechev
With LLM watermarking being deployed commercially and now required by regulations, improving its reliability and effectiveness has become crucial. Yet, recent progress in the field…
Marking the Wrong Symptoms: Evaluating LLM Watermarks in Medical Texts
Melanie Rieff, Robin Staab, Thibaud Gloaguen +2
Large language models (LLMs) are increasingly integrated into clinical workflows, stressing the need for reliable traceability of model-generated output with watermarking. Yet, mos…
Making Open-Source Text LLM Watermarks Durable Against Merging
Luisa Scharff, Thibaud Gloaguen, Robin Staab +1
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…
Every Bit, Everywhere, All at Once: A Binomial Multibit LLM Watermark
Thibaud Gloaguen, Robin Staab, Mark Vero +1
With LLM watermarking already being deployed commercially, practical applications increasingly require multibit watermarks that encode more complex payloads, such as user IDs or ti…
Coding Agents Don't Know When to Act
Thibaud Gloaguen, Niels Mündler, Mark Müller +2
Coding agents are increasingly deployed to autonomously maintain software, including to resolve user-reported issues: a bug report comes in and the agent creates a patch to address…
A Unified Framework for LLM Watermarks
Thibaud Gloaguen, Robin Staab, Nikola Jovanović +1
LLM watermarks allow tracing AI-generated texts by inserting a detectable signal into their generated content. Recent works have proposed a wide range of watermarking algorithms, e…
Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?
Thibaud Gloaguen, Niels Mündler-Sasahara, Niels Mündler +4
A widespread practice in software development is to tailor coding agents to repositories using context files, such as AGENTS.md. Although this practice is strongly encouraged by ag…
Fewer Weights, More Problems: A Practical Attack on LLM Pruning
Kazuki Egashira, Robin Staab, Thibaud Gloaguen +2
Model pruning, i.e., removing a subset of model weights, has become a prominent approach to reducing the memory footprint of large language models (LLMs) during inference. Notably,…
Watermarking Diffusion Language Models
Thibaud Gloaguen, Robin Staab, Nikola Jovanović +1
We introduce the first watermark tailored for diffusion language models (DLMs), an emergent LLM paradigm able to generate tokens in arbitrary order, in contrast to standard autoreg…
LLM Fingerprinting via Semantically Conditioned Watermarks
Thibaud Gloaguen, Robin Staab, Nikola Jovanović +1
Most LLM fingerprinting methods teach the model to respond to a few fixed queries with predefined atypical responses (keys). This memorization often does not survive common deploym…
Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM Finetuning
Thibaud Gloaguen, Mark Vero, Robin Staab +1
Finetuning open-weight Large Language Models (LLMs) is standard practice for achieving task-specific performance improvements. Until now, finetuning has been regarded as a controll…
Towards Watermarking of Open-Source LLMs
Thibaud Gloaguen, Nikola Jovanović, Robin Staab +1
While watermarks for closed LLMs have matured and have been included in large-scale deployments, these methods are not applicable to open-source models, which allow users full cont…
Discovering Spoofing Attempts on Language Model Watermarks
Thibaud Gloaguen, Nikola Jovanović, Robin Staab +1
LLM watermarks stand out as a promising way to attribute ownership of LLM-generated text. One threat to watermark credibility comes from spoofing attacks, where an unauthorized thi…
Black-Box Detection of Language Model Watermarks
Thibaud Gloaguen, Nikola Jovanović, Robin Staab +1
Watermarking has emerged as a promising way to detect LLM-generated text, by augmenting LLM generations with later detectable signals. Recent work has proposed multiple families of…