AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Bayesian Low-Rank Posteriors for Scalable Membership Inference

Khanh Nguyen, Antonio Cinà, Andrey Barsky, Alejandra Reinares, Fabio Roli, Dimosthenis Karatzas

NeurIPS 2026 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Nonparametric and PAC-Bayes Bayesian learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes a scalable membership inference attack for large VLMs that replaces explicit reference/shadow model ensembles with a Bayesian low-rank posterior approximation.

Problem

Membership inference attacks based on reference/shadow model ensembles are computationally prohibitive for large-scale vision-language models.

Approach

Replaces explicit reference models with a Bayesian low-rank posterior that estimates target-conditional score distributions, avoiding training many shadow models.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Replace reference-model ensemble with a SWAG posterior over LoRA adapters (single fine-tuning run)LIBRA on DeepSeek-VL2 · DocVQA multi-sample MIA, D_realTPR@3%FPR: 14.0% vs 5.3% strongest baselineImproved
Image-only setting: LIBRA vs strongest baselineIdefics2 · DocVQA image-only MIA, D_realTPR@3%FPR: 33.0% vs 11.3%Improved
Use K=15 posterior samples instead of K=4 (ablation)LIBRA Qwen2-VL · DocVQA multi-sample (validation)TPR@3%FPR: 27.18 -> 31.67Improved
At matched ensemble size K=N=4, LIBRA vs LiRA (ablation)Qwen2-VL · DocVQA multi-sample (validation)TPR@3%FPR / AUROC: 22.17 vs 27.18 TPR; 79.42 vs 77.09 AUROCNo clear effect
Use SWAG posterior instead of MC Dropout (ablation)LIBRA · DocVQA multi-sample MIA, D_realTPR@3%FPR: DeepSeek-VL2 6.33 -> 14.00; Qwen2-VL 42.00 -> 38.33Mixed
Reference adapter mismatch (LoRA reference vs VeRA target) (ablation)LIBRA vs LiRA/RMIA · Cross-adapter transfer, D_realTPR@3%FPR: LIBRA 27.33 vs LiRA 20.67 vs RMIA 18.67Mixed
Increase posterior scale factor kappa from 2 to 8 (ablation)LIBRA Qwen2-VL · Per-sample calibrationobserved FPR: 12.67% -> 9%Improved
Increase inference temperature beta of target model as a defense (ablation)reference-based MIAs · Qwen2-VL DocVQAattack strengthHurt
Model reference scores with population-based quantile regression (Q-MIA) rather than reference modelsQwen2-VL DocVQA · Per-sample calibrated MIA alpha=0.03TPR: 7.67 vs LIBRA 61.67Hurt
Use one SWAG-based reference model instead of N=4 trained shadow modelsQwen2-VL · Training costtraining GPU-hours: 5.66 -> 1.30Improved
Apply LIBRA to sample-level Medical VQAQwen2-VL · SLAKE sample-level MIATPR@3%FPR: 3.55 -> 5.46Improved

Experimental setup

Builds on: reference-model membership inference attacks

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy