SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.04808 · 2024

WaterMax: breaking the LLM watermark detectability-robustness-quality trade-off

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 3 functions out of this paper's own repositories and ran 1 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
eva-giboulot/watermax canonical 1 of 3
FunctionStatusWhere it lives
generate_mmw_path Ran eva-giboulot/watermax/test_sentence_wm.py
pointer only (licence: NONE) · get_code("2006c2f6f6daa764")
get_detector Not yet run eva-giboulot/watermax/test_sentence_wm.py
pointer only (licence: NONE) · get_code("aa5abf5d87851c1c")
get_generator Not yet run eva-giboulot/watermax/test_sentence_wm.py
pointer only (licence: NONE) · get_code("f59da376d411ca93")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Watermarking is a technical means to dissuade malfeasant usage of Large Language Models. This paper proposes a novel watermarking scheme, so-called WaterMax, that enjoys high detectability while sustaining the quality of the generated text of the original LLM. Its new design leaves the LLM untouched (no modification of the weights, logits, temperature, or sampling technique). WaterMax balances robustness and complexity contrary to the watermarking techniques of the literature inherently provoking a trade-off between quality and robustness. Its performance is both theoretically proven and experimentally validated. It outperforms all the SotA techniques under the most complete benchmark suite. Code available at https://github.com/eva-giboulot/WaterMax.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2403.04808")
get_code_for_paper("2403.04808")
have("2403.04808")

Connect an agent — have() is free.