Language Model Fingerprinting Requires Rethinking Watermark Teachers

Excerpt

arXiv:2610.04169v1 Announce Type: cross Abstract: LLM fingerprinting via watermark distillation embeds a statistical watermark signal into model weights, enabling model owners to identify their models behind black-box APIs. Revisiting a recent protocol, we find that its utility evaluation understates text quality degradation in open-ended generation, favoring overly strong watermark teachers. Weakening the watermark improves text quality but sacrifices detectability. To move beyond this trade-of