mlsanityData to 7 Oct 2026

Have You Merged My Model? On The Robustness of Large Language Model IP Protection Methods Against Model Merging

arXiv 2404.05188 ↗First posted 8 Apr 2024cs.CR cs.AI cs.CLcited by 17

On the map: LLM safety and security → Watermarking generated content and models

Abstract (arXiv)

Model merging is a promising lightweight model empowerment technique that does not rely on expensive computing devices (e.g., GPUs) or require the collection of specific training data. Instead, it involves editing different upstream model parameters to absorb their downstream task capabilities. However, uncertified model merging can infringe upon the Intellectual Property (IP) rights of the original upstream models. In this paper, we conduct the first study on the robustness of IP protection methods under model merging scenarios. Specifically, we investigate two state-of-the-art IP protection techniques: Quantization Watermarking and Instructional Fingerprint, along with various advanced model merging technologies, such as Task Arithmetic, TIES-MERGING, and so on. Experimental results indicate that current Large Language Model (LLM) watermarking techniques cannot survive in the merged models, whereas model fingerprinting techniques can. Our research aims to highlight that model merging should be an indispensable consideration in the robustness assessment of model IP protection techniques, thereby promoting the healthy development of the open-source LLM community. Our code is available at https://github.com/ThuCCSLab/MergeGuard.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Placed in Watermarking generated content and models, one of 12 directions in LLM safety and security. Builds on 18 of its 26 references that are in the map. First posted on arXiv 8 Apr 2024.

Builds on

18 of 26 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023Watermarking LLMs with Weight Quantization17 Oct
  2. 2023Provable Robust Watermarking for AI-Generated Text30 Jun

Other references · 16

  1. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  2. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  3. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
  4. 2022Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference timecited by 1,366
  5. 2022Editing Models with Task Arithmeticcited by 1,076
  6. 2023TIES-Merging: Resolving Interference When Merging Modelscited by 795
  7. 2023Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunchcited by 679
  8. 2023WizardMath: Empowering Mathematical Reasoning for Large Language Models via Reinforced Evol-Instructcited by 618
  9. 2018Turning Your Weakness Into a Strength: Watermarking Deep Neural Networks by Backdooringcited by 334
  10. 2023Are aligned neural networks adversarially aligned?cited by 305
  11. 2024Arcee's MergeKit: A Toolkit for Merging Large Language Modelscited by 189
  12. 2024Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmeticcited by 105
4 more
  1. 2019IPGuard: Protecting Intellectual Property of Deep Neural Networks via Fingerprinting the Classification Boundarycited by 68
  2. 2024Instructional Fingerprinting of Large Language Modelscited by 64
  3. 2022SSLGuard: A Watermarking Scheme for Self-supervised Learning Pre-trained Encoderscited by 43
  4. 2024Here's a Free Lunch: Sanitizing Backdoored Models with Model Mergecited by 21

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 4; 2025: 8; 2026: 5.

17 papers.

Sort:

By area:

  1. 2024Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunitiescited by 247
  2. 2024Mitigating the Backdoor Effect for Multi-Task Model Merging via Safety-Aware Subspacecited by 14
  3. 2025Scalable Fingerprinting of Large Language Modelscited by 11
  4. 2025Disrupting Model Merging: A Parameter-Level Defense Without Sacrificing Accuracycited by 11
  5. 2024Fingerprint Vector: Enabling Scalable and Efficient Model Fingerprint Transfer via Vector Additioncited by 9
  6. 2025MEraser: An Effective Fingerprint Erasure Approach for Large Language Modelscited by 9
  7. 2024HM3: Hierarchical Multi-Objective Model Merging for Pretrained Modelscited by 8
  8. 2025Model Unmerging: Making Your Models Unmergeable for Secure Model Sharingcited by 6
  9. 2025A Systematic Study of In-the-Wild Model Merging for Large Language Modelscited by 4
  10. 2025RouteMark: A Fingerprint for Intellectual Property Attribution in Routing-based Model Mergingcited by 3
  11. 2025Do Not Merge My Model! Safeguarding Open-Source LLMs Against Unauthorized Model Mergingcited by 2
  12. 2025Defending Unauthorized Model Merging via Dual-Stage Weight Protectioncited by 2
5 more
  1. 2026ForgetMark: Stealthy Fingerprint Embedding via Targeted Unlearning in Language Modelscited by 1
  2. 2026Making Models Unmergeable via Scaling-Sensitive Loss Landscapecited by 1
  3. 2026From Parameters to Behaviors: A Survey of Model Fusion for Large Language Modelscited by 1
  4. 2026DNF: Dual-Layer Nested Fingerprinting for Large Language Model Intellectual Property Protectioncited by 0
  5. 2026On the Vulnerability of Parameter-Level Defenses to Model Mergingcited by 0