Watermarkability: Measuring and Improving Language-Model Compatibility with Logit-Based Watermarks

Key points

• Language-model watermarks can help identify model-generated text, encode provenance information, and detect post-generation edits.1
• A broad class of logit-based watermarks increases the logits of a designated token set at each decoding step.1
• At a target mass $τ$, we quantify this property through a watermarkability cost: the expected minimum KL divergence from the base-model distribution required to assign probability at least $τ$ to the requested set.1
• Across the 12 model-schedule pairs, the pre-adaptation watermarkability cost correlates with the log-PPL increase under a near-hard fixed-bias baseline (pooled Pearson $r=0.929$).1

Sources (1)

• [1]Watermarkability: Measuring and Improving Language-Model Compatibility with Logit-Based Watermarks
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 01:40 AM

  We introduce watermarkability to characterize the compatibility between a model and the token sets requested by a watermark.

  Language-model watermarks can help identify model-generated text, encode provenance information, and detect post-generation edits.

Extractive summary: sentences quoted from the sources.