Improves safety region certification for smoothed classifiers without changing smoothing scheme.
problem Certified safety regions for smoothed classifiers are often small compared to optimal.
method Generalizes certified radius calculation as nested optimization problem, uses 0th-1st order information, and designs efficient estimators.
result Certified safety regions are significantly larger than current methods, achieving significant improvements on various metrics.
Paper aims to ensure reliable detection of out-of-distribution data with certifiable worst-case guarantees.
problem Deep neural networks are overconfident with OOD inputs, posing safety risks.
method Enforces low confidence and bounds in an l∞-ball around OOD points using interval bound propagation (IBP). result Certifiable worst-case guarantees for OOD detection are possible without significant loss in accuracy.
BagCert efficiently certifies robustness against adversarial patches on image classifiers.
problem Adversarial patches pose a threat to autonomous systems' perception component.
method BagCert combines model architecture and certification procedure for efficient inference.
result BagCert certifies 10,000 examples in 43 seconds on a single GPU, achieving 86% clean and 60% certified accuracy against 5x5 patches.
Training neural networks to be certifiably robust is critical to ensure their safety against adversarial attacks. However, it is currently very difficult to train a neural network that is both accurate and certifiably robust. In this work we take a step towards addressing this challenge. We prove that for every continu…
The paper certifies neural network-based control barrier functions efficiently.
problem Certifying neural network-based barrier functions for safety in autonomous systems.
method Combines NN reachability and hyperplane arrangement enumeration for efficient certification.
result Soundly finds regions where neural networks are certified as barrier functions.
CIVET method provides robustness guarantees for VAEs under adversarial attacks.
problem Certified probabilistic guarantees for VAEs in safety-critical applications.
method Bounding worst-case VAE error by error on support sets at the latent layer.
result CIVET outperforms state-of-the-art methods in robustness and standard performance.
Study probabilistic safety of BNNs under adversarial attacks.
problem Evaluate vulnerability of BNNs to adversarial attacks.
method Relaxation techniques from non-convex optimization to compute probabilistic safety bounds.
result Certify probabilistic safety of BNNs with millions of parameters.
New method certifies neural network robustness to random input noise.
problem Certifying neural networks' robustness to random input noise.
method Chance-constrained optimization problem reformulated using input-output samples.
result Certifies a uniform infinity-norm uncertainty region with a radius 50 times larger.
This paper provides a comprehensive benchmark and taxonomy for certifiably robust DNN defenses.
problem Certifiably robust defenses against adversarial attacks for deep neural networks.
method Taxonomy and benchmark of certifiably robust approaches.
result First comprehensive benchmark of certifiably robust approaches on different datasets.
New method certifies neural network robustness under random input noise.
problem Certifying neural network robustness against random input noise.
method Chance-constrained optimization problem reformulated with input-output samples, convex conditions developed.
result Proposed method certifies robustness against various input noise regimes over larger uncertainty regions.
Framework for safely updating machine learning models.
problem Continuous updates to machine learning models can lead to unintended consequences.
method Formalizes the problem as computing the largest locally invariant domain (LID), uses tractable primal-dual formulation.
result Matches or exceeds heuristic baselines for avoiding forgetting while providing formal safety guarantees.
Efficient local Lipschitz bounds improve neural network robustness.
problem Certifying robustness of neural networks is challenging and often leads to over-regularization.
method Proposes an efficient trainable local Lipschitz upper bound by considering activation functions and weight matrices.
result Consistently outperforms state-of-the-art methods in clean and certified accuracy on various datasets.
Certified calibration methods protect model confidence from adversarial attacks.
problem Adversarial attacks degrade model calibration, reducing confidence in predictions.
method Developed certified calibration methods to provide worst-case bounds on calibration under adversarial perturbations.
result Certified calibration methods produce analytic and approximate bounds for the Brier score and expected calibration error.
GP3 framework efficiently analyzes Gaussian processes on GPUs.
problem Certifiable safety in machine learning applications.
method GP3 framework using interval analysis and multi-resolution sampling on GPUs.
result Efficient analysis of Gaussian processes with certifiable safety.
Deep-PrAE improves rare-event simulation for black-box systems.
problem Evaluating rare safety-critical events in learning-based systems.
method Combines deep neural networks with IS to create statistically guaranteed estimations.
result Deep-PrAE provides accurate bounds on safety-critical event probabilities.
Paper reviews robustness in machine learning models and discusses training and certification methods.
problem Ensuring reliability of machine learning models in safety-critical systems.
method Reviews formalisms and discusses training and certification techniques.
result Identifies future research directions in robust machine learning.
Novel neural network approximates exact distance for robust classification.
problem Adversarial attacks on neural networks in safety-critical systems.
method Signed Distance Classifiers (SDCs) and Unitary-Gradient Neural Network.
result Approximates exact distance from classification boundary for certifiable predictions.
Develops first robustness verification for complex Transformers.
problem Certify prediction behavior of Transformers with complex self-attention layers.
method Resolves challenges of cross-nonlinearity and cross-position dependency in Transformers.
result Certified robustness bounds are significantly tighter than Interval Bound Propagation.
This work improves deep reinforcement learning robustness to adversarial state uncertainty.
problem Robustness of deep reinforcement learning to adversarial state uncertainty.
method Certified adversarial robustness techniques are applied to deep reinforcement learning algorithms to compute guaranteed lower bounds on state-action values.
result The approach increases robustness to noise and adversaries in pedestrian collision avoidance and classic control tasks.
Proposes a method to ensure accurate estimation of rare events in AI systems.
problem Lack of efficiency guarantees in black-box systems for rare-event simulation.
method Integrates deep learning with importance sampling to create a statistically guaranteed estimator.
result Demonstrates effective estimation of rare-event probabilities in AI systems.
ViTaX provides formal guarantees for targeted explanations in safety-critical systems.
problem Need trustworthy explanations for safety-critical deep neural networks.
method Formal reachability analysis for targeted, semifactual explanations.
result First method to provide formally guaranteed explanations of model resilience.
The paper provides guarantees for feedback control with sensor errors.
problem Certifying performance and safety in feedback control systems with sensor errors.
method Solving a supervised learning problem to characterize sensor errors and providing uniform error bounds.
result Finite-time convergence rate on sub-optimality of using a regressor in closed-loop for waypoint tracking.
Reinforcement Learning (RL) is a widely employed machine learning architecture that has been applied to a variety of control problems. However, applications in safety-critical domains require a systematic and formal approach to specifying requirements as tasks or goals. We propose a model-free RL algorithm that enables…
Fusion of robustness and uncertainty techniques improves adversarial defense.
problem Adversarial attacks on deep neural networks.
method Integrating uncertainty quantification into randomized smoothing for robustness guarantees.
result Improved robustness guarantees for uncertainty aware classifiers.
New method certifies regression robustness without data distribution assumptions.
problem Certifying robustness of regression models against poisoning attacks.
method Reduces certified regression to certified classification using median decision function.
result Proposes six new provably-robust regression models.
Two impossibility theorems show formal alignment certification is impossible for AI systems.
problem Formal certification of AI alignment over open-ended domains is impossible.
method Two independent impossibility theorems: Semantic and Statistical barriers.
result No procedure can simultaneously satisfy soundness, completeness, and tractability.
Proposes new convex relaxations for certifying spatial robustness of neural networks.
problem Lack of provable guarantees for robustness against vector field transformations.
method Novel convex relaxations for certifying robustness against vector field transformations.
result First time providing a certificate of robustness against vector field transformations.
SingleProp speeds up robust neural network training with minimal certification.
problem Efficiently defending neural networks against adversarial attacks with certified guarantees.
method SingleProp regularizer that requires only one forward pass per training iteration.
result Comparable certified accuracy to state-of-the-art defenses, but significantly faster training.
New attack tricks certifiably robust models into mislabeling images.
problem Defeating certified defenses against adversarial examples.
method Spoofed robustness certificates and large perturbations.
result Certifiably robust models can be fooled by large perturbations.
SAFER method certifies robustness to word substitutions without model structure.
problem Certified robustness against synonymous word substitutions in NLP models.
method Randomized smoothing with stochastic ensemble of randomized inputs.
result Significantly outperforms state-of-the-art methods for certified robustness.
New techniques extend certified unlearning to deep neural networks.
problem Applying certified unlearning to deep neural networks (DNNs) is challenging due to their nonconvex nature.
method Developed simple techniques and an efficient computation method for nonconvex objectives, considering nonconvergence training and sequential unlearning.
result Demonstrated the efficacy of the method on real-world datasets, showing advantages of certified unlearning in DNNs.
This work improves certifiably robust models by distilling knowledge from adversarially robust teachers.
problem Certifiably robust models suffer from poor standard performance.
method Knowledge distillation from adversarially robust teachers to improve standard performance.
result Distillation from adversarially robust teachers consistently improves certified training performance.
It is well-known that classifiers are vulnerable to adversarial perturbations. To defend against adversarial perturbations, various certified robustness results have been derived. However, existing certified robustnesses are limited to top-1 predictions. In many real-world applications, top-k predictions are more rel…
One epoch training yields certifiably robust models.
problem Vulnerability of machine learning models to adversarial attacks.
method Deterministic certification approach based on regularized loss.
result Certifiable robust models achieved in one epoch.
Paper presents certified defenses against adversarial patch attacks.
problem Certified defenses against adversarial patch attacks are needed.
method Proposes the first certified defense and faster training methods.
result Demonstrates robustness transfer across different patch shapes.
New algorithms learn stability certificates from data, avoiding complex dynamics.
problem Synthesizing stability certificates from complex dynamical systems.
method Developed algorithms to learn certificate functions from trajectory data, establishing generalization error bounds.
result Efficiently learned certificates can be used for adaptive control.
New method improves certified robustness for classifier confidence.
problem Certifying confidence in classifier predictions.
method Randomized smoothing with modified Neyman-Pearson lemma.
result Certified radii for prediction confidence improved.
New regularizers tighten convex relaxation bounds for neural networks.
problem Large gap between certifiable and empirical robustness in neural networks.
method Two regularizers to train neural networks yielding tighter convex relaxation bounds.
result Higher certified accuracy with proposed regularizers.
Certified training improves robustness against adversarial attacks.
problem Certified training's gap with empirical robustness limits its practical utility.
method Combining adversarial attacks with network over-approximations.
result Certified training can prevent catastrophic overfitting and bridge the gap to multi-step baselines.
This paper proposes a framework for certifying neural network defenses against data poisoning attacks.
problem Vulnerability of neural networks to data poisoning attacks.
method Random selection based defenses that average predictions on sub-datasets sampled from the training set.
result The certified radius of bagging derived by the framework is tighter than previous work.
Improved defense against data poisoning attacks by aggregating smaller subsets.
problem Mitigating the impact of poisoned data on model robustness.
method Finite Aggregation method that combines duplicates of smaller disjoint subsets for training.
result Consistent improvement in certified robustness bounds, up to 4.77% on GTSRB.
New work shows limits of certifying neural network robustness.
problem Certified training improves robustness but decreases accuracy.
method Bayes error analysis to investigate robustness limits.
result Upper bound for certified robust accuracy established.
New approach detects adversarial samples with certifiable guarantees.
problem Adversarial samples can trick CNNs, posing a threat.
method Certifiable Taboo Trap (CTT) approach to detect adversarial inputs.
result CTT outperforms existing defenses on various lp norms. Paper improves neural network robustness certification with tighter radii estimates.
problem Certifying neural networks' robustness against adversarial attacks.
method Advanced algorithms for discrete and continuous domains, optimizing sample size, standard deviation, and temperature.
result Significant improvement in certified test-set accuracy with tighter certified radii bounds.
We show how to turn any classifier that classifies well under Gaussian noise into a new classifier that is certifiably robust to adversarial perturbations under the ℓ2 norm. This "randomized smoothing" technique has been proposed recently in the literature, but existing guarantees are loose. We prove a tight robu…
New method certifies risks of LLM outputs, improving accuracy and reliability.
problem Uncertain and incorrect outputs from large language models.
method Information-lift certificates using PAC-Bayes bounds and skeleton design.
result Achieves 77.0% coverage at 2% risk, outperforming baselines.
New framework tightens certified robustness gaps in machine learning models.
problem Persistent gap between theoretical certified robustness and empirical accuracy.
method Leverages Lipschitz continuity and novel confidence intervals.
result Improves robust accuracy, compressing the gap between theory and practice.
Paper proposes faster certified robust training methods with short warmup.
problem Certified robust training methods require long warmup schedules, making training costly.
method Proposes three improvements: new weight initialization, BN, and regularization.
result Achieves 65.03% verified error on CIFAR-10 and 82.36% on TinyImageNet with short warmup.