Edit Distance Robust Watermarks via Indexing Pseudorandom Codes

Authors: Noah Golowich, Ankur Moitra

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical As this paper is theoretical in nature, we work only with this abstraction of an autoregressive model
Researcher Affiliation Academia Noah Golowich nzg@mit.edu MIT Ankur Moitra moitra@mit.edu MIT
Pseudocode Yes Algorithm 1 Indexing PRC: PRCIdx[PRCSub, ρ], Algorithm 2 PRF-PRC[F, p, q]: Generic PRF-based PRC, Algorithm 3 Watermarking from PRCs for general alphabets: W[PRC, Model]
Open Source Code No The paper is theoretical and does not mention releasing any source code.
Open Datasets No The paper is theoretical and does not use or mention any specific training datasets.
Dataset Splits No The paper is theoretical and does not involve dataset splits (training, validation, test).
Hardware Specification No The paper is theoretical and does not describe any hardware used for experiments.
Software Dependencies No The paper is theoretical and does not mention specific software dependencies with version numbers.
Experiment Setup No The paper is theoretical and does not describe any experimental setup, hyperparameters, or training configurations.