PACZero: PAC-Private Fine-Tuning of Language Models via Sign Quantization

We introduce PACZero, a family of PAC-private zeroth-order mechanisms for fine-tuning large language models that delivers usable utility at $I(S^*; Y_{1:T})=0$. This privacy regime bounds the membership-inference attack (MIA) posterior success rate at the prior, an MIA-resistance level the DP framework matches only at $\varepsilon=0$ and infinite noise. All DP-ZO comparisons below are matched at the MIA posterior level. The key insight is that PAC Privacy charges mutual information only when the release depends on which candidate subset is the secret. Sign-quantizing subset-aggregated zeroth-order gradients creates frequent unanimity, steps at which every candidate subset agrees on the update direction; at these steps the released sign costs zero conditional mutual information. We propose two variants that span the privacy-utility trade-off: PACZero-MI (budgeted MI via exact calibration on the binary release) and PACZero-ZPL ($I=0$ via a uniform coin flip on disagreement steps). We evaluate on SST-2 and SQuAD with OPT-1.3B and OPT-6.7B in both LoRA and full-parameter tracks. On SST-2 OPT-1.3B full fine-tuning at $I=0$, PACZero-ZPL reaches ${88.99\pm0.91}$, within $2.1$pp of the non-private MeZO baseline ($91.1$ FT). No prior method produces usable utility in the high-privacy regime $\varepsilon<1$, and PACZero-ZPL obtains competitive SST-2 accuracy and nontrivial SQuAD F1 across OPT-1.3B and OPT-6.7B at $I=0$.

Trade-off Functions for DP-SGD with Subsampling based on Random Shuffling: Tight Upper and Lower Bounds

We derive a tight analysis of the trade-off function for Differentially Private Stochastic Gradient Descent (DP-SGD) with subsampling based on random shuffling within the $f$-DP framework. Our analysis covers the regime $σ\geq \sqrt{3/\ln M}$, where $σ$ is the noise multiplier and $M$ is the number of rounds within a single epoch. Unlike $f$-DP analyses for Poisson subsampling, which yield non-closed implicit formulas that can be machine computed but are non-transparent, random shuffling admits a tight analysis yielding transparent and interpretable closed-form bounds. Our concrete bounds, derived via the Berry-Esseen theorem, are tight up to constant factors within the proof framework. We demonstrate worked parameter settings for a single epoch ($E=1$) with a corresponding trade-off function $\geq 1-a-δ$, that is, only $δ$ below the ideal random guessing diagonal $1-a$: For $δ= 1/100$ and $σ= 1$, roughly $M \approx 1.14\times 10^6$ rounds and $N \approx 1.14\times 10^7$ training samples suffice to achieve meaningful differential privacy. This is in contrast to recent negative results for the regime $σ\leq 1/\sqrt{2 \ln M}$. Our concrete bounds can be composed over multiple epochs leading to $δ$ having a linear in $E$ dependency, which restricts $E=O(\sqrt{M})$. To go beyond Berry--Esseen, we introduce a new proof technique based on a generalization of the law of large numbers that yields an asymptotic random guessing diagonal-limit result: if $E=c_M^2M$ with $c_M\to 0$, then the $E$-fold composed trade-off function satisfies $f^{\otimes E}(a)\to 1-a$ uniformly in $a\in[0,1]$ with $δ$ having only an $O(\sqrt{E})$ dependency. We compare this asymptotic regime with the corresponding Poisson subsampling asymptotic, and highlight the characterization of explicit convergence rates as an open question.

TOSSS: a CVE-based Software Security Benchmark for Large Language Models

With their increasing capabilities, Large Language Models (LLMs) are now used across many industries. They have become useful tools for software engineers and support a wide range of development tasks. As LLMs are increasingly used in software development workflows, a critical question arises: are LLMs good at software security? At the same time, organizations worldwide invest heavily in cybersecurity to reduce exposure to disruptive attacks. The integration of LLMs into software engineering workflows may introduce new vulnerabilities and weaken existing security efforts. We introduce TOSSS (Two-Option Secure Snippet Selection), a benchmark that measures the ability of LLMs to choose between secure and vulnerable code snippets. Existing security benchmarks for LLMs cover only a limited range of vulnerabilities. In contrast, TOSSS relies on the CVE database and provides an extensible framework that can integrate newly disclosed vulnerabilities over time. Our benchmark gives each model a security score between 0 and 1 based on its behavior; a score of 1 indicates that the model always selects the secure snippet, while a score of 0 indicates that it always selects the vulnerable one. We evaluate 14 widely used open-source and closed-source models on C/C++ and Java code and observe scores ranging from 0.48 to 0.89. LLM providers already publish many benchmark scores for their models, and TOSSS could become a complementary security-focused score to include in these reports.

On the Evidentiary Limits of Membership Inference for Copyright Auditing

As large language models (LLMs) are trained on increasingly opaque corpora, membership inference attacks (MIAs) have been proposed to audit whether copyrighted texts were used during training, despite growing concerns about their reliability under realistic conditions. We ask whether MIAs can serve as admissible evidence in adversarial copyright disputes where an accused model developer may obfuscate training data while preserving semantic content, and formalize this setting through a judge-prosecutor-accused communication protocol. To test robustness under this protocol, we introduce SAGE (Structure-Aware SAE-Guided Extraction), a paraphrasing framework guided by Sparse Autoencoders (SAEs) that rewrites training data to alter lexical structure while preserving semantic content and downstream utility. Our experiments show that state-of-the-art MIAs degrade when models are fine-tuned on SAGE-generated paraphrases, indicating that their signals are not robust to semantics-preserving transformations. While some leakage remains in certain fine-tuning regimes, these results suggest that MIAs are brittle in adversarial settings and insufficient, on their own, as a standalone mechanism for copyright auditing of LLMs.

Fundamental Limitations of Favorable Privacy-Utility Guarantees for DP-SGD

Differentially Private Stochastic Gradient Descent (DP-SGD) is the dominant paradigm for private training, but its fundamental limitations under worst-case adversarial privacy definitions remain poorly understood. We analyze DP-SGD in the $f$-differential privacy framework, which characterizes privacy via hypothesis-testing trade-off curves, and study shuffled sampling over a single epoch with $M$ gradient updates. We derive an explicit suboptimal upper bound on the achievable trade-off curve. This result induces a geometric lower bound on the separation $κ$ which is the maximum distance between the mechanism's trade-off curve and the ideal random-guessing line. Because a large separation implies significant adversarial advantage, meaningful privacy requires small $κ$. However, we prove that enforcing a small separation imposes a strict lower bound on the Gaussian noise multiplier $σ$, which directly limits the achievable utility. In particular, under the standard worst-case adversarial model, shuffled DP-SGD must satisfy $σ\ge \frac{1}{\sqrt{2\ln M}}$ $\quad\text{or}\quad$ $κ\ge\ \frac{1}{\sqrt{8}}\!\left(1-\frac{1}{\sqrt{4π\ln M}}\right)$, and thus cannot simultaneously achieve strong privacy and high utility. Although this bound vanishes asymptotically as $M \to \infty$, the convergence is extremely slow: even for practically relevant numbers of updates the required noise magnitude remains substantial. We further show that the same limitation extends to Poisson subsampling up to constant factors. Our experiments confirm that the noise levels implied by this bound leads to significant accuracy degradation at realistic training settings, thus showing a critical bottleneck in DP-SGD under standard worst-case adversarial assumptions.

Beyond Anonymization: Object Scrubbing for Privacy-Preserving 2D and 3D Vision Tasks

We introduce ROAR (Robust Object Removal and Re-annotation), a scalable framework for privacy-preserving dataset obfuscation that eliminates sensitive objects instead of modifying them. Our method integrates instance segmentation with generative inpainting to remove identifiable entities while preserving scene integrity. Extensive evaluations on 2D COCO-based object detection show that ROAR achieves 87.5% of the baseline detection average precision (AP), whereas image dropping achieves only 74.2% of the baseline AP, highlighting the advantage of scrubbing in preserving dataset utility. The degradation is even more severe for small objects due to occlusion and loss of fine-grained details. Furthermore, in NeRF-based 3D reconstruction, our method incurs a PSNR loss of at most 1.66 dB while maintaining SSIM and improving LPIPS, demonstrating superior perceptual quality. Our findings establish object removal as an effective privacy framework, achieving strong privacy guarantees with minimal performance trade-offs. The results highlight key challenges in generative inpainting, occlusion-robust segmentation, and task-specific scrubbing, setting the foundation for future advancements in privacy-preserving vision systems.