# Fine-Tuning Integrity for Modern Neural Networks: Structured Drift Proofs via Norm, Rank, and Sparsity Certificates

arXiv: https://arxiv.org/abs/2604.04738 · HTML: https://mlsanity.com/paper/2604.04738/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 41 references that are in the map.
- First posted on arXiv 6 Apr 2026.

## Builds on (most cited first, 12 of 12)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Parameter-Efficient Transfer Learning for NLP](https://mlsanity.com/paper/1902.00751/) (2019)
- [Parameter-Efficient Fine-Tuning for Large Models: A Comprehensive Survey](https://mlsanity.com/paper/2403.14608/) (2024)
- [Parameter-Efficient Fine-Tuning Methods for Pretrained Language Models: A Critical Review and Assessment](https://mlsanity.com/paper/2312.12148/) (2023)
- [CrypTFlow: Secure TensorFlow Inference](https://mlsanity.com/paper/1909.07814/) (2019)
- [What Makes and Breaks Safety Fine-tuning? A Mechanistic Study](https://mlsanity.com/paper/2407.10264/) (2024)
- [Chiron: Privacy-preserving Machine Learning as a Service](https://mlsanity.com/paper/1803.05961/) (2018)
- [PEFTGuard: Detecting Backdoor Attacks Against Parameter-Efficient Fine-Tuning](https://mlsanity.com/paper/2411.17453/) (2024)
- [A Survey of Zero-Knowledge Proof Based Verifiable Machine Learning](https://mlsanity.com/paper/2502.18535/) (2025)
- [Time-Optimal Interactive Proofs for Circuit Evaluation](https://mlsanity.com/paper/1304.3812/) (2013)
- [A Study of Backdoors in Instruction Fine-tuned Language Models](https://mlsanity.com/paper/2406.07778/) (2024)
- [A Survey on the Applications of Zero-Knowledge Proofs](https://mlsanity.com/paper/2408.00243/) (2024)

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

- [ServeGuard: Verifiable, Bounded-Residual Confinement of Operator-Invisible Channels Without Revealing the Certified Read Factor](https://mlsanity.com/paper/2609.21515/) (2026)

## Abstract (arXiv)

Fine-tuning is the dominant paradigm for adapting large machine learning models, yet current deployment pipelines provide no way to verify how a released model was updated. In particular, a model provider or auditor cannot check whether a fine-tuned model adheres to a claimed update procedure without access to its parameters. We introduce fine-tuning integrity (FTI), a cryptographic objective for verifying that a deployed model differs from a trusted base model only within a declared class of admissible updates. We construct succinct model difference proofs (SMDPs), zero-knowledge protocols that certify structured parameter drift without revealing model weights. Our framework supports three fundamental update classes: norm-bounded, low-rank, and sparse drift, covering common fine-tuning methods such as regularized training, LoRA, and prefix tuning. In all cases, proof size and verification cost depend on the structure of the update rather than the number of parameters. We prove soundness, zero-knowledge, and succinctness for each construction, and establish a matching Omega(n) lower bound showing that structural assumptions are necessary for succinct verification. A prototype evaluation on synthetic benchmarks and GPT-2 fine-tuning demonstrates that proofs remain compact and verification is efficient at realistic scales.