# WaterMoE: Expert-Routing-based Watermarking for High Fidelity and Efficiency

arXiv: https://arxiv.org/abs/2607.13099 · HTML: https://mlsanity.com/paper/2607.13099/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Expert routing strategies, one of 12 directions in Mixture-of-experts.
- Builds on 40 of its 46 references that are in the map.
- First posted on arXiv 14 Jul 2026.

## Placed in

- Mixture-of-experts / [Expert routing strategies](https://mlsanity.com/moe/expert-routing-strategies/)

## Builds on (most cited first, 30 of 40)

- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Mixtral of Experts](https://mlsanity.com/paper/2401.04088/) (2024)
- [Competition-Level Code Generation with AlphaCode](https://mlsanity.com/paper/2203.07814/) (2022)
- [Instruction-Following Evaluation for Large Language Models](https://mlsanity.com/paper/2311.07911/) (2023)
- [Measuring Coding Challenge Competence With APPS](https://mlsanity.com/paper/2105.09938/) (2021)
- [A Watermark for Large Language Models](https://mlsanity.com/paper/2301.10226/) (2023)
- [A Survey on Large Language Models for Code Generation](https://mlsanity.com/paper/2406.00515/) (2024)
- [ELI5: Long Form Question Answering](https://mlsanity.com/paper/1907.09190/) (2019)
- [Multi-News: a Large-Scale Multi-Document Summarization Dataset and Abstractive Hierarchical Model](https://mlsanity.com/paper/1906.01749/) (2019)
- [Understanding the planning of LLM agents: A survey](https://mlsanity.com/paper/2402.02716/) (2024)
- [DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale](https://mlsanity.com/paper/2201.05596/) (2022)
- [Large Language Models for Information Retrieval: A Survey](https://mlsanity.com/paper/2308.07107/) (2023)
- [Robust Distortion-free Watermarks for Language Models](https://mlsanity.com/paper/2307.15593/) (2023)
- [BookSum: A Collection of Datasets for Long-form Narrative Summarization](https://mlsanity.com/paper/2105.08209/) (2021)
- [LLaMA-MoE: Building Mixture-of-Experts from LLaMA with Continual Pre-training](https://mlsanity.com/paper/2406.16554/) (2024)
- [A Semantic Invariant Robust Watermark for Large Language Models](https://mlsanity.com/paper/2310.06356/) (2023)
- [Unbiased Watermark for Large Language Models](https://mlsanity.com/paper/2310.10669/) (2023)
- [Uni-MoE: Scaling Unified Multimodal LLMs with Mixture of Experts](https://mlsanity.com/paper/2405.11273/) (2024)
- [WritingBench: A Comprehensive Benchmark for Generative Writing](https://mlsanity.com/paper/2503.05244/) (2025)
- [MarkLLM: An Open-Source Toolkit for LLM Watermarking](https://mlsanity.com/paper/2405.10051/) (2024)
- [A Survey of LLM-based Agents in Medicine: How far are we from Baymax?](https://mlsanity.com/paper/2502.11211/) (2025)
- [An Entropy-based Text Watermarking Detection Method](https://mlsanity.com/paper/2403.13485/) (2024)
- [An Unforgeable Publicly Verifiable Watermark for Large Language Models](https://mlsanity.com/paper/2307.16230/) (2023)
- [LLM Agents for Education: Advances and Applications](https://mlsanity.com/paper/2503.11733/) (2025)
- [k-SemStamp: A Clustering-Based Semantic Watermark for Detection of Machine-Generated Text](https://mlsanity.com/paper/2402.11399/) (2024)
- [Can Watermarks Survive Translation? On the Cross-lingual Consistency of Text Watermark for Large Language Models](https://mlsanity.com/paper/2402.14007/) (2024)
- [A Continued Pretrained LLM Approach for Automatic Medical Note Generation](https://mlsanity.com/paper/2403.09057/) (2024)

## Built on by (most cited first, 1 of 1)

- [Beyond Semantic Narrowing: Robust and Efficient LLM Watermarking with Hamming Neighborhoods](https://mlsanity.com/paper/2609.37218/) (2026)

## Abstract (arXiv)

Large language models (LLMs) have achieved remarkable success but raise growing concerns about content provenance and misuse, motivating the need for reliable watermarking techniques. However, these techniques have rarely been adopted in practice mainly for two reasons: i) severely degraded model performance, and ii) additional inference overhead. To confirm the problem, we construct a comprehensive benchmark spanning different generation tasks to systematically evaluate 9 representative watermarking methods. We found almost all existing methods are designed for text fluency, but not for restricted and complicated tasks, and their overhead prevents them from deployment in latency-critical systems. To address i) and ii), we propose an LLM watermarking scheme textitWaterMoE for the growingly popular Mixture-of-Experts (MoE) LLMs. WaterMoE embeds watermarking signals through controlled perturbation into the expert selection at each router, which accumulates to token selection shift at the final output. In contrast to watermarking as a post-processing token-sampling approach, WaterMoE embeds watermark within the inference loop incurring negligible quality degradation and computational overhead. Extensive experiments demonstrate that our method achieves a fidelity performance close to the unwatermarked and consistently outperforms state-of-the-art watermarking methods on the benchmark, with up to 4× speedup, incurring merely 1% additional inference latency compared to native generation. The results demonstrate the capability of WaterMoE to be deployed in real-world tasks.