Overview

Certified robustness studies whether a model’s predictions or actions remain safe within a defined range of perturbations, rather than relying only on empirical attack tests. Common approaches include interval bound propagation, convex relaxations and branch-and-bound verification, Lipschitz bounds, and randomized smoothing; exact methods such as mixed-integer programming can provide stronger guarantees on smaller problems. Researchers apply these tools to image and language models, transformers, time-series systems, reinforcement-learning agents, and safety-critical applications, using benchmarks such as MNIST, CIFAR, and ImageNet as well as domain-specific tests. A central challenge is balancing tight, realistic certificates with the computation needed to obtain them, especially when perturbations involve semantic changes, physical conditions, multiple input types, or finite-precision execution.

Latest papers

99 papers10 papers in the last four weeks, down 29% on the four weeks before. 0.1% of all new papers.

Feb 2, 2026

Learning Better Certified Models from Empirically-Robust Teachers

Adversarial training attains strong empirical robustness to specific adversarial attacks by training on concrete adversarial perturbations, but it produces neural networks that are not amenable to strong robustness certificates through neural network verification. On the other hand, earlier certified training schemes directly train on bounds from network relaxations to obtain models that are certifiably robust, but display sub-par standard performance. Recent work has shown that state-of-the-art trade-offs between certified robustness and standard performance can be obtained through a family of losses combining adversarial outputs and neural network bounds. Nevertheless, differently from empirical robustness, verifiability still comes at a significant cost in standard performance. In this work, we propose to leverage empirically-robust teachers to improve the performance of certifiably-robust models through knowledge distillation. Using versatile feature-space distillation objectives, we show that distillation from adversarially-trained teachers can significantly improve the performance of state-of-the-art certified training algorithms for ReLU networks on robust computer vision benchmarks.
Jan 22, 2026

Feature-Space Smoothing: Certified Robustness of Deep Representations

Modern deep learning models exhibit strong capabilities across diverse applications, yet remain vulnerable to malicious inputs that induce erroneous predictions via feature-space distortion. To address that, we propose Feature-Space Smoothing (FS), a representation-level certification framework built upon Gaussian randomized smoothing. We show that FS converts a given feature encoder into a smoothed variant that is guaranteed to maintain a certified lower bound on the cosine similarity between clean and adversarial features under l_2-bounded perturbations. We then establish that this Feature Cosine Similarity Bound (FCSB) can be extended to the prediction-wise certification under the cosine similarity measure, and the value of FCSB is determined by the encoder intrinsic Gaussian robustness score. Building on those insights, we introduce the Gaussian Smoothness Booster (GSB), a plug-and-play module to improve the encoder Gaussian robustness score. Specifically, the GSB module is plugged to enhance the feature-space consistency and maintain the feature utility for downstream tasks under Gaussian perturbations. This design enables seamless integration of FS on the protected model, e.g., Multimodal Large Language Models (MLLMs), without additional model retraining or alignment, improving its robustness while preserving the performance for downstream task-oriented decoding. Extensive experiments demonstrate that integrating FS consistently provides non-trivial certified robustness and significantly improves task-oriented performance under strong white-box adversarial attacks across diverse models and applications.
Dec 23, 2025

Pareto DNN Verification: Fast for Most Queries

Formal verification of deep neural networks (DNNs) faces a fundamental scalability challenge: verifying robustness is NP-hard and state-of-the-art tools often time out on modern architectures. We propose PAreto VErification, or PAVE: by augmenting DNNs with early exits (EEs) and verifying exit-by-exit, we solve a large fraction of robustness queries---roughly 80% in our evaluation---in a small fraction of the time required for direct verification. The key insight is that many inputs in a robustness neighborhood exit at an early layer. Verifying this early-exit behavior is often a tractable problem,allowing verification to terminate long before the full network is analyzed. EEs can be added without modifying the base network, and incur at most a 0.3% accuracy change across our tested architectures. We formalize the novel robustness property for EE networks, prove it is Fixed-Parameter Tractable under a trace stability condition (which commonly holds in practice), and present a sound and complete verification algorithm. Two sound optimizations --- break (early termination when the winner dominates) and continue (skipping the inner loop when no runner-up can win) --- reduce SAFE-case verification time by up to 10x. Experiments on MNIST, CIFAR-10, and CIFAR-100 with networks from 600K to 33M parameters demonstrate that PAVE enables solving queries that were not solved otherwise, including formal verification of ResNet-18 on CIFAR-100.
Dec 1, 2025

CluCERT: Certifying LLM Robustness via Clustering-Guided Denoising Smoothing

Recent advancements in Large Language Models (LLMs) have led to their widespread adoption in daily applications. Despite their impressive capabilities, they remain vulnerable to adversarial attacks, as even minor meaning-preserving changes such as synonym substitutions can lead to incorrect predictions. As a result, certifying the robustness of LLMs against such adversarial prompts is of vital importance. Existing approaches focused on word deletion or simple denoising strategies to achieve robustness certification. However, these methods face two critical limitations: (1) they yield loose robustness bounds due to the lack of semantic validation for perturbed outputs and (2) they suffer from high computational costs due to repeated sampling. To address these limitations, we propose CluCERT, a novel framework for certifying LLM robustness via clustering-guided denoising smoothing. Specifically, to achieve tighter certified bounds, we introduce a semantic clustering filter that reduces noisy samples and retains meaningful perturbations, supported by theoretical analysis. Furthermore, we enhance computational efficiency through two mechanisms: a refine module that extracts core semantics, and a fast synonym substitution strategy that accelerates the denoising process. Finally, we conduct extensive experiments on various downstream tasks and jailbreak defense scenarios. Experimental results demonstrate that our method outperforms existing certified approaches in both robustness bounds and computational efficiency.
Dec 1, 2025

Dual Randomized Smoothing: Beyond Global Noise Variance

Randomized Smoothing (RS) is a prominent technique for certifying the robustness of neural networks against adversarial perturbations. With RS, achieving high accuracy at small radii requires a small noise variance, while achieving high accuracy at large radii requires a large noise variance. However, the global noise variance used in the standard RS formulation leads to a fundamental limitation: there exists no global noise variance that simultaneously achieves strong performance at both small and large radii. To break through the global variance limitation, we propose a dual RS framework which enables input-dependent noise variances. To achieve that, we first prove that RS remains valid with input-dependent noise variances, provided the variance is locally constant around each input. Building on this result, we introduce two components: (i) a variance estimator predicts an optimal noise variance for each input, (ii) this estimated variance is then used by a standard RS classifier. The variance estimator is independently smoothed via RS to ensure local constancy, enabling flexible design. We also introduce training strategies to iteratively optimize the two components. Experiments on CIFAR-10 demonstrate that our dual RS method provides strong performance for both small and large radii-unattainable with global noise variance-while incurring only a 60% computational overhead at inference. Moreover, it outperforms prior input-dependent noise approaches across most radii, with gains at radii 0.5, 0.75, and 1.0 of 15.6%, 20.0%, and 15.7%. On ImageNet, dual RS remains effective across all radii, with advantages of 8.6%, 17.1%, and 9.1% at radii 0.5, 1.0, and 1.5. Additionally, the dual RS framework provides a routing perspective for certified robustness, improving the accuracy-robustness trade-off with off-the-shelf expert RS models.
Jan 31, 2025

Towards the Worst-case Robustness of Large Language Models

Recent studies have revealed the vulnerability of large language models to adversarial attacks, where adversaries craft specific inputs to induce incorrect or harmful outputs. Although various empirical defenses have been proposed, their worst-case robustness remains poorly understood, raising concerns about how they would behave against stronger future attacks. Here, we study the worst-case robustness of LLMs from both empirical and theoretical perspectives. First, we provide tighter empirical upper bounds for the tested deterministic defenses using stronger white-box attacks, under which most tested configurations achieve nearly 0% robustness. Second, we formulate randomized-smoothing certification for bounded and binary base functions as fractional and 0-1 knapsack problems, respectively, and derive grouped solvers for two token kernels. For example, on AdvBench, we certify a GPT-4o safety detector smoothed by a uniform kernel to an average certified ℓ0\ell_0 radius of 2.02. In our stated suffix-attack setting, the same kernel certifies a Llama-3-8B safety detector to an average certified radius of 6.41.
Oct 9, 2024

Expressiveness of Multi-Neuron Convex Relaxations in Neural Network Certification

Neural network certification methods heavily rely on convex relaxations to provide robustness guarantees. However, these relaxations are often imprecise: even the most accurate single-neuron relaxation is incomplete for general ReLU networks, a limitation known as the single-neuron convex barrier. While multi-neuron relaxations have been heuristically applied to address this issue, two central questions arise: (i) whether they overcome the convex barrier, and if not, (ii) whether they offer theoretical capabilities beyond those of single-neuron relaxations. In this work, we present the first rigorous analysis of the expressiveness of multi-neuron relaxations. Perhaps surprisingly, we show that they are inherently incomplete, even when allocated sufficient resources to capture finitely many neurons and layers optimally. This result extends the single-neuron barrier to a universal convex barrier for neural network certification. On the positive side, we show that completeness can be achieved by either (i) augmenting the network with a polynomial number of carefully designed ReLU neurons or (ii) partitioning the input domain into convex sub-polytopes, thereby distinguishing multi-neuron relaxations from single-neuron ones which are unable to realize the former and have worse partition complexity for the latter. Our findings establish a foundation for multi-neuron relaxations and point to new directions for certified robustness, including training methods tailored to multi-neuron relaxations and verification methods with multi-neuron relaxations as the main subroutine.
Aug 17, 2024

Training Verifiably Robust Agents Using Set-Based Reinforcement Learning

Reinforcement learning policies parametrized by deep neural networks have achieved strong performance for continuous control, yet even small input perturbations may lead to unpredictable behavior. This sensitivity limits their use in safety-critical domains, where robustness guarantees are required. Our work addresses this gap between state-of-the-art adversarial training methods and formal verification to train verifiably robust agents. Previous works train networks with individual adversarial perturbations, making them only robust against the specific adversarial attacks used. In contrast, our approach propagates entire perturbed input sets, enclosing all possible adversarial attacks within a single network pass. We leverage this to explicitly penalize the size of the output set (minimizing closed-loop uncertainty) and thereby make the actor robust against all possible attacks. This is realized by the use of set-based policy gradients, where each output within the set has a different gradient, thereby balancing the accuracy and robustness of the network. Doing so, we achieve formal verifiability across different verification frameworks for up to 9 times larger input perturbations compared to standard reinforcement learning and improve certified worst-case performance.
Jun 17, 2023

Understanding Certified Training with Interval Bound Propagation

As robustness verification methods are becoming more precise, training certifiably robust neural networks is becoming ever more relevant. To this end, certified training methods compute and then optimize an upper bound on the worst-case loss over a robustness specification. Curiously, training methods based on the imprecise interval bound propagation (IBP) consistently outperform those leveraging more precise bounding methods. Still, we lack an understanding of the mechanisms making IBP so successful. In this work, we thoroughly investigate these mechanisms by leveraging a novel metric measuring the tightness of IBP bounds. We first show theoretically that, for deep linear models, tightness decreases with width and depth at initialization, but improves with IBP training, given sufficient network width. We, then, derive sufficient and necessary conditions on weight matrices for IBP bounds to become exact and demonstrate that these impose strong regularization, explaining the empirically observed trade-off between robustness and accuracy in certified training. Our extensive experimental evaluation validates our theoretical predictions for ReLU networks, including that wider networks improve performance, yielding state-of-the-art results. Interestingly, we observe that while all IBP-based training methods lead to high tightness, this is neither sufficient nor necessary to achieve high certifiable robustness. This hints at the existence of new training methods that do not induce the strong regularization required for tight IBP bounds, leading to improved robustness and standard accuracy.