mlsanityData to 7 Oct 2026

Provable Model Provenance Set for Large Language Models

arXiv 2602.00772 ↗First posted 31 Jan 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

The growing prevalence of unauthorized model usage and misattribution has increased the need for reliable model provenance analysis. However, existing methods largely rely on heuristic fingerprint-matching rules that lack provable error control and often overlook the existence of multiple sources, leaving the reliability of their provenance claims unverified. In this work, we first formalize the model provenance problem with provable guarantees, requiring rigorous coverage of all true provenances at a prescribed confidence level. Then, we propose the Model Provenance Set (MPS), which employs a sequential test-and-exclusion procedure to adaptively construct a small set satisfying the guarantee. The key idea of MPS is to test the significance of provenance existence within a candidate pool, thereby establishing a provable asymptotic guarantee at a user-specific confidence level. Extensive experiments demonstrate that MPS effectively achieves target provenance coverage while strictly limiting the inclusion of unrelated models, and further reveal its potential for practical provenance analysis in attribution and auditing tasks.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 26 of its 33 references that are in the map. First posted on arXiv 31 Jan 2026.

Builds on

26 of 33 references.

References · 26

  1. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
  2. 2019HuggingFace's Transformers: State-of-the-art Natural Language Processingcited by 2,201
  3. 2024MiniCPM-V: A GPT-4V Level MLLM on Your Phonecited by 1,061
  4. 2021A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantificationcited by 777
  5. 2023A Watermark for Large Language Modelscited by 667
  6. 2023A Survey on Large Language Model (LLM) Security and Privacy: The Good, the Bad, and the Uglycited by 380
  7. 2023Robust Distortion-free Watermarks for Language Modelscited by 279
  8. 2023A Semantic Invariant Robust Watermark for Large Language Modelscited by 97
  9. 2023REMARK-LLM: A Robust and Efficient Watermarking Framework for Generative Large Language Modelscited by 64
  10. 2023Are You Copying My Model? Protecting the Copyright of Large Language Models for EaaS via Backdoor Watermarkcited by 61
  11. 2024LLMmap: Fingerprinting For Large Language Modelscited by 56
  12. 2024TRAP: Targeted Random Adversarial Prompt Honeypot for Black-Box Identificationcited by 33
14 more
  1. 2023Watermarking LLMs with Weight Quantizationcited by 19
  2. 2025Copyright Protection for Large Language Models: A Survey of Methods, Challenges, and Trendscited by 19
  3. 2025Intrinsic Fingerprint of LLMs: Continue Training is NOT All You Need to Steal A Model!cited by 16
  4. 2025Model Provenance Testing for Large Language Modelscited by 15
  5. 2025LLM DNA: Tracing Model Evolution via Functional Representationscited by 15
  6. 2025Gradient-Based Model Fingerprinting for LLM Similarity Detection and Family Classificationcited by 13
  7. 2025CTCC: A Robust and Stealthy Fingerprinting Framework for Large Language Models via Cross-Turn Contextual Correlation Backdoorcited by 11
  8. 2025Blackbox Model Provenance via Palimpsestic Membership Inferencecited by 11
  9. 2025Fingerprinting LLMs via Prompt Injectioncited by 8
  10. 2023Turning Your Strength into Watermark: Watermarking Large Language Model via Knowledge Injectioncited by 7
  11. 2025Are Robust LLM Fingerprints Adversarially Robust?cited by 7
  12. 2025Knowledge Distillation Detection for Open-weights Modelscited by 4
  13. 2025WaterMod: Modular Token-Rank Partitioning for Probability-Balanced LLM Watermarkingcited by 4
  14. 2025SELF: A Robust Singular Value and Eigenvalue Approach for LLM Fingerprintingcited by 2

Built on by

Not cited yet.