Search papers, labs, and topics across Lattice.
6
0
6
0
A groundbreaking watermarking method that not only traces LLM-generated content but also detects tampering with unprecedented accuracy.
Watermarks that are intrinsically tied to image semantics can effectively thwart forgery attempts, a breakthrough in the fight against image manipulation.
A watermark can not only identify machine-generated text but also attribute it to specific users and extract hidden information, all governed by a precise information-theoretic framework.
Watermarking LLM-agent trajectories just got a major upgrade鈥擳RACE achieves near-perfect detection without sacrificing performance, even under aggressive adversarial conditions.
Diffusion watermarks, thought to be robust, crumble under a simple stochastic resampling attack that breaks trajectory reconstruction.
LLMs can cleverly rewrite generative images to erase semantic watermarks, even when those watermarks are designed to be robust to content-aware modifications.
