WorldCup Sampling for Multi-bit LLM Watermarking

TL;DR

This paper introduces WorldCup, a novel multi-bit watermarking framework for LLMs that models sampling as a communication channel, improving robustness, quality, and capacity over prior methods.

Contribution

WorldCup models LLM sampling as a structured communication channel and uses hierarchical competition and entropy-aware modulation for improved watermarking.

Findings

Outperforms prior methods in capacity, robustness, and text quality.

Balances message detectability and decoding efficiency.

Provides a scalable, principled approach for multi-bit watermarking.

Abstract

As large language models (LLMs) generate increasingly human-like text, watermarking has emerged as a promising solution for reliable attribution beyond mere detection. While multi-bit watermarking enables richer provenance encoding, existing approaches typically extend zero-bit watermarking schemes by introducing static logit perturbations and counting-based decoding strategies, which can degrade text quality and compromise decoding robustness as the payload increases. In this paper, we propose WorldCup, a multi-bit watermarking framework for LLMs that models the sampling process as a structured communication channel and embeds message bits through a hierarchical competition mechanism guided by complementary signals. Moreover, WorldCup incorporates entropy-aware modulation to preserve generation quality and enables robust message recovery via confidence-aware decoding that accounts for token-level reliability. Comprehensive experiments demonstrate that WorldCup achieves a strong balance across message capacity, detectability, robustness, text quality, and decoding efficiency, consistently outperforming prior baselines. We believe that this work establishes a scalable and principled foundation for future research on multi-bit watermarking in LLMs.