Conference paper · 2026 · Open access

In-Context Probing for Membership Inference in Fine-Tuned Language Models

Zhengyu Lu, Hongliang Chi, Nathalie Baracaldo, Swanand Kadhe, Yuseok Jeon, Lei Yu

Network and Distributed System Security Symposium · Published 1 January 2026 · DOI 10.14722/ndss.2026.240892

Open PDFCite this paper

In brief

“In-Context Probing for Membership Inference in Fine-Tuned Language Models” is a 2026 conference paper by Zhengyu Lu and 5 co-authors, published in Network and Distributed System Security Symposium. OpenAlex counts 7 citations as of 19 November 2026, which puts it in the top 2% of works published in 2026 by citations. It is open access: the full text is free to read at Network and Distributed System Security Symposium.

Citations
7
Most cited in
2026 (7)
Field-weighted impact
6.20
References
64
Authors
6

Abstract

membership inference [5], [6].Among these stages, finetuning is particularly susceptible to privacy leaks, due to the typically limited size and sensitive nature of the fine-tuning datasets [7].Membership inference attacks (MIAs) aim to determine whether a particular data sample was part of a model's training dataset, thereby potentially revealing sensitive or personally identifiable information about individuals [8].MIAs have been extensively studied across a range of machine learning domains to identify and characterize privacy risks.These include generative adversarial networks (GANs) [9], explainable machine learning models [10], and diffusion models [11].In addition to vulnerability assessment, MIAs have also been employed to evaluate the efficacy of privacy-preserving mechanisms [12], [13], benchmark machine unlearning methods [14], and enable privacy auditing in deployed systems [15]-[17].Recently, a growing body of work [18], [19] has adapted MIAs to assess the privacy risks of LLMs.These efforts build on classical MIA techniques but tailor them to the unique properties of LLMs.Broadly, existing MIA methods for LLMs can be categorized into two classes: reference-based attacks, which rely on an auxiliary/reference dataset (typically drawn from a distribution similar to the model's training data) to train one or more reference models, and referencefree attacks, which avoid this requirement.Reference-based attacks, such as those by Mireshghallah et al. [15], [18], extend the Likelihood Ratio Attack (LiRA) framework [20] to LLMs and masked language models.These attacks require training an ensemble of shadow models and comparing the target model's negative log-likelihood (NLL) on a given sample against the distribution of NLLs from these shadow models to determine the membership.However, such approaches assume access to auxiliary data from the same distribution as the target model's training set, an assumption that rarely holds in real-world scenarios, especially when fine-tuning involves private or proprietary datasets.In contrast, reference-free attacks such as Min-K% [5] and Min-K%++ [21] detect memorized samples by identifying low-rank (outlier) tokens in the model's output, which are indicative of overfitting.ReCaLL [22] demonstrates that adding a context prefix to the input can differentially affect the model's predictions for memorized versus non-memorized Abstract-Membership inference attacks (MIAs) pose a critical privacy threat to fine-tuned large language models (LLMs), especially when models are adapted to domain-specific tasks using sensitive data.While prior black-box MIA techniques rely on confidence scores or token likelihoods, these signals are often entangled with a sample's intrinsic properties-such as content difficulty or rarity-leading to poor generalization and low signal-to-noise ratios.In this paper, we propose ICP-MIA, a novel MIA framework grounded in the theory of training dynamics, particularly the phenomenon of diminishing returns during optimization.We introduce the Optimization Gap as…

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Lu, Z., Chi, H., Baracaldo, N., Kadhe, S., Jeon, Y., & Yu, L. (2026). In-Context Probing for Membership Inference in Fine-Tuned Language Models. Network and Distributed System Security Symposium. https://doi.org/10.14722/ndss.2026.240892

MLA 9

Lu, Zhengyu, et al. “In-Context Probing for Membership Inference in Fine-Tuned Language Models.” Network and Distributed System Security Symposium, 2026. https://doi.org/10.14722/ndss.2026.240892.

Chicago

Lu, Zhengyu, Hongliang Chi, Nathalie Baracaldo, Swanand Kadhe, Yuseok Jeon, and Lei Yu. 2026. “In-Context Probing for Membership Inference in Fine-Tuned Language Models.” Network and Distributed System Security Symposium. https://doi.org/10.14722/ndss.2026.240892.

Harvard

Lu, Z. et al. (2026) ‘In-Context Probing for Membership Inference in Fine-Tuned Language Models’, Network and Distributed System Security Symposium. doi:10.14722/ndss.2026.240892.

IEEE

Z. Lu, H. Chi, N. Baracaldo, S. Kadhe, Y. Jeon, and L. Yu, “In-Context Probing for Membership Inference in Fine-Tuned Language Models,” Network and Distributed System Security Symposium, 2026, doi: 10.14722/ndss.2026.240892.

BibTeX

@inproceedings{lu2026context,
  title     = {{In-Context Probing for Membership Inference in Fine-Tuned Language Models}},
  author    = {Lu, Zhengyu and Chi, Hongliang and Baracaldo, Nathalie and Kadhe, Swanand and Jeon, Yuseok and Yu, Lei},
  booktitle = {Network and Distributed System Security Symposium},
  year      = {2026},
  doi       = {10.14722/ndss.2026.240892},
  url       = {https://doi.org/10.14722/ndss.2026.240892}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data

  • membership inference attacks
  • fine-tuned large language models
  • memorization
  • privacy leakage
  • shadow model
  • overfitting

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Lu, Z., Chi, H., Baracaldo, N., Kadhe, S., Jeon, Y., & Yu, L. (2026). In-Context Probing for Membership Inference in Fine-Tuned Language Models. Network and Distributed System Security Symposium. https://doi.org/10.14722/ndss.2026.240892

How many times has this paper been cited?

7 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Network and Distributed System Security Symposium: sign in and open the PDF from this page.

Who wrote this paper?

Zhengyu Lu, Hongliang Chi, Nathalie Baracaldo, Swanand Kadhe, Yuseok Jeon, Lei Yu.

What is the DOI of this paper?

10.14722/ndss.2026.240892, permanently at https://doi.org/10.14722/ndss.2026.240892. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.