Responsibly deploying artificial intelligence (AI) / machine learning (ML) systems in high-stakes settings arguably requires not only proof of system reliability, but also continual, post-deployment monitoring to quickly detect and address any unsafe behavior. Methods for nonparametric sequential testing—especially conformal test martingales (CTMs) and anytime-valid inference—offer promising tools for this monitoring task. However, existing approaches are restricted to monitoring limited hypothesis classes or “alarm criteria” (e.g., detecting data shifts that violate certain exchangeability or IID assumptions), do not allow for online adaptation in response to shifts, and/or cannot diagnose the cause of degradation or alarm. In this paper, we address these limitations by proposing a weighted generalization of conformal test martingales (WCTMs), which lay a theoretical foundation for online monitoring for any unexpected changepoints in the data distribution while controlling false-alarms. For practical applications, we propose specific WCTM algorithms that adapt online to mild covariate shifts (in the marginal input distribution), quickly detect harmful shifts, and diagnose those harmful shifts as concept shifts (in the conditional label distribution) or extreme (out-of-support) covariate shifts that cannot be easily adapted to. On real-world datasets, we demonstrate improved performance relative to state-of-the-art baselines.

Token-inconsistency bugs (TIBs) involve the misuse of syntactically valid yet incorrect code tokens, such as misused variables and erroneous function invocations, which can often lead to software bugs. Unlike simple syntactic bugs, TIBs occur at the semantic level and are subtle - sometimes they remain undetected for years. Traditional detection methods, such as static analysis and dynamic testing, often struggle with TIBs due to their versatile and context-dependent nature. However, advancements in large language models (LLMs) like GPT-4 present new opportunities for automating TIB detection by leveraging these models’ semantic understanding capabilities.This paper reports the first systematic measurement of LLMs’ capabilities in detecting TIBs, revealing that while GPT-4 shows promise, it exhibits limitations in precision and scalability. Specifically, its detection capability is undermined by the model’s tendency to focus on the code snippets that do not contain TIBs; its scalability concern arises from GPT-4’s high cost and the massive amount of code requiring inspection. To address these challenges, we introduce LineBreaker, a novel and cascaded TIB detection system. LineBreaker leverages smaller, codespecific, and highly efficient language models to filter out large numbers of code snippets unlikely to contain TIBs, thereby significantly enhancing the system’s performance in terms of precision, recall, and scalability. We evaluated LineBreaker on 154 Python and C GitHub repositories, each with over 1,000 stars, uncovering 123 new flaws, 45% of which could be exploited to disrupt program functionalities. Out of our 69 submitted fixes, 41 have already been confirmed or merged.

Confidential computing (CC), designed for security-critical scenarios, uses remote attestation to guarantee code integrity on cloud servers. However, CC alone cannot provide assurance of high-level security properties (e.g., no data leak) on the code. In this paper, we introduce a novel framework, Agora , scrupulously designed to provide a trustworthy and open verification platform for CC. To prompt trustworthiness, we observe that certain verification tasks can be delegated to untrusted entities, while the corresponding (smaller) validators are securely housed within the trusted computing base (TCB). Moreover, through a novel blockchain-based bounty task manager, it also utilizes crowdsourcing to remove trust in complex theorem provers. These synergistic techniques successfully ameliorate the TCB size burden associated with two procedures: binary analysis and theorem proving. To prompt openness, Agora supports a versatile assertion language that allows verification of various security policies. Moreover, the design of Agora enables untrusted parties to participate in any complex processes out of Agora ’s TCB. By implementing verification workflows for software-based fault isolation, information flow control, and side-channel mitigation policies, our evaluation demonstrates the efficacy of Agora .

As machine learning models in critical fields increasingly grapple with multimodal data, they face the dual challenges of handling a wide array of modalities, often incomplete due to missing elements, and the temporal irregularity and sparsity of collected samples. Successfully leveraging this complex data, while overcoming the scarcity of high-quality training samples, is key to improving these models' predictive performance. We introduce ``FuseMoE'', a mixture-of-experts framework incorporated with an innovative gating function. Designed to integrate a diverse number of modalities, FuseMoE is effective in managing scenarios with missing modalities and irregularly sampled data trajectories. Theoretically, our unique gating function contributes to enhanced convergence rates, leading to better performance in multiple downstream tasks. The practical utility of FuseMoE in the real world is validated by a diverse set of challenging prediction tasks.

We propose a new, principled approach to tackling missing data problems that can reduce both bias and variance of any (stochastic) gradient descent-based predictive model that is learned on such data. The proposed method can use an arbitrary (and potentially biased) imputation model to fill in the missing values, as it corrects the biases introduced by imputation with a control variates method, leading to an unbiased estimation for gradient updates. Theoretically, we prove that our control variates approach improves the convergence of stochastic gradient descent under common missing data settings. Empirically, we show that our method yields superior performance as compared to the results obtained using competing imputation methods, on various applications, across different missing data patterns.

Transformer-based architectures have recently exhibited remarkable successes across different domains beyond just powering large language models. However, existing approaches typically focus on predictive accuracy and computational cost, largely ignoring certain other practical issues such as robustness to contaminated samples. In this paper, by re-interpreting the self-attention mechanism as a non-parametric kernel density estimator, we adapt classical robust kernel density estimation methods to develop novel classes of transformers that are resistant to adversarial attacks and data contamination. We first propose methods that down-weight outliers in RKHS when computing the self-attention operations. We empirically show that these methods produce improved performance over existing state-of-the-art methods, particularly on image data under adversarial attacks. Then we leverage the median-of-means principle to obtain another efficient approach that results in noticeably enhanced performance and robustness on language modeling and time series classification tasks. Our methods can be combined with existing transformers to augment their robust properties, thus promising to impact a wide variety of applications.

A recently emerged cellular network based One-Tap Authentication (OTAuth) scheme allows app users to quickly sign up or log in to their accounts conveniently: Mobile Network Operator (MNO) provided tokens instead of user passwords are used as identity credentials. After conducting a first in-depth security analysis, however, we have revealed several fundamental design flaws among popular OTAuth services, which allow an adversary to easily (1) perform unauthorized login and register new accounts as the victim, (2) illegally obtain identities of victims, and (3) interfere OTAuth services of legitimate apps. To further evaluate the impact of our identified issues, we propose a pipeline that integrates both static and dynamic analysis. We examined 1,025/894 Android/iOS apps, each app holding more than 100 million installations. We confirmed 396/398 Android/iOS apps are affected. Our research systematically reveals the threats against OTAuth services. Finally, we provide suggestions on how to mitigate these threats accordingly.

With growing concerns regarding data privacy and rapid increase in data volume, Federated Learning (FL) has become an important learning paradigm. However, jointly learning a deep neural network model in a FL setting proves to be a non-trivial task because of the complexities associated with the neural networks, such as varied architectures across clients, permutation invariance of the neurons, and presence of non-linear transformations in each layer. This work introduces a novel framework, Federated Heterogeneous Neural Networks (FedHeNN), that allows each client to build a personalised model without enforcing a common architecture across clients. This allows each client to optimize with respect to local data and compute constraints, while still benefiting from the learnings of other (potentially more powerful) clients. The key idea of FedHeNN is to use the instance-level representations obtained from peer clients to guide the simultaneous training on each client. The extensive experimental results demonstrate that the FedHeNN framework is capable of learning better performing models on clients in both the settings of homogeneous and heterogeneous architectures across clients.

While considering the spatial and temporal features of traffic, capturing the impacts of various external factors on travel is an essential step towards achieving accurate traffic forecasting. However, existing studies seldom consider external factors or neglect the effect of the complex correlations among external factors on traffic. Intuitively, knowledge graphs can naturally describe these correlations. Since knowledge graphs and traffic networks are essentially heterogeneous networks, it is challenging to integrate the information in both networks. On this background, this study presents a knowledge representation-driven traffic forecasting method based on spatial-temporal graph convolutional networks. We first construct a knowledge graph for traffic forecasting and derive knowledge representations by a knowledge representation learning method named KR-EAR. Then, we propose the Knowledge Fusion Cell (KF-Cell) to combine the knowledge and traffic features as the input of a spatial-temporal graph convolutional backbone network. Experimental results on the real-world dataset show that our strategy enhances the forecasting performances of backbones at various prediction horizons. The ablation and perturbation analysis further verify the effectiveness and robustness of the proposed method. To the best of our knowledge, this is the first study that constructs and utilizes a knowledge graph to facilitate traffic forecasting; it also offers a promising direction to integrate external information and spatial-temporal information for traffic forecasting. The source code is available athttps://github.com/lehaifeng/T-GCN/tree/master/KST-GCN.

Many real-life applications involve simultaneously forecasting multiple time series that are hierarchically related via aggregation or disaggregation operations. For instance, commercial organizations often want to forecast inventories simultaneously at store, city, and state levels for resource planning purposes. In such applications, it is important that the forecasts, in addition to being reasonably accurate, are also consistent w.r.t one another. Although forecasting such hierarchical time series has been pursued by economists and data scientists, the current state-of-the-art models use strong assumptions, e.g., all forecasts being unbiased estimates, noise distribution being Gaussian. Besides, state-of-the-art models have not harnessed the power of modern nonlinear models, especially ones based on deep learning. In this paper, we propose using a flexible nonlinear model that optimizes quantile regression loss coupled with suitable regularization terms to maintain the consistency of forecasts across hierarchies. The theoretical framework introduced herein can be applied to any forecasting model with an underlying differentiable loss function. A proof of optimality of our proposed method is also provided. Simulation studies over a range of datasets highlight the efficacy of our approach.

How can we find a subset of training samples that are most responsible for a specific prediction made by a complex black-box machine learning model? More generally, how can we explain the model's decisions to end-users in a transparent way? We propose a new model-agnostic algorithm to identify a minimal set of training samples that are indispensable for a given model's decision at a particular test point, i.e., the model's decision would have changed upon the removal of this subset from the training dataset. Our algorithm identifies such a set of “indispensable” samples iteratively by solving a constrained optimization problem. Further, we speed up the algorithm through efficient approximations and provide theoretical justification for its performance. To demonstrate the applicability and effectiveness of our approach, we apply it to a variety of tasks including data poisoning detection, training set debugging and understanding loan decisions. The results show that our algorithm is an effective and easy-to-comprehend tool that helps to better understand local model behavior, and therefore facilitates the adoption of machine learning in domains where such understanding is a requisite.

Learning monotonic models with respect to a subset of the inputs is a desirable feature to effectively address the fairness, interpretability, and generalization issues in practice. Existing methods for learning monotonic neural networks either require specifically designed model structures to ensure monotonicity, which can be too restrictive/complicated, or enforce monotonicity by adjusting the learning process, which cannot provably guarantee the learned model is monotonic on selected features. In this work, we propose to certify the monotonicity of the general piece-wise linear neural networks by solving a mixed integer linear programming problem. This provides a new general approach for learning monotonic neural networks with arbitrary model structures. Our method allows us to train neural networks with heuristic monotonicity regularizations, and we can gradually increase the regularization magnitude until the learned network is certified monotonic. Compared to prior work, our method does not require human-designed constraints on the weight space and also yields more accurate approximation. Empirical studies on various datasets demonstrate the efficiency of our approach over the state-of-the-art methods, such as Deep Lattice Networks

Sequences of 221 alpha-helical antimicrobial peptides (alphaAMPs) were compared and 63-166 of them were selected and analyzed using Perl programs. The results showed that aliphatic amino acids Gly, Leu, Ala, Ile and two positively charged amino acids Lys and Arg were composed of more than 63% of the first 20 residues of alphaAMPs. The weighed mean membrane partitioning energies at positions from 1 to 25 of alphaAMPs were calculated. Profile of the partitioning energies suggests oblique membrane insertion and an amphipathic alpha-helical structure of the N-terminus of alphaAMP (residues from 1 to 13), a bend structure at positions 13 and 14, and a less structured C-terminus that parallels the surface of the membrane. These structural features are in good agreement with the experimentally determined membrane structure of hemagglutinin fusion peptide from influenza virus. We hypothesize that this (N-terminal oblique alpha-helix)-central bend-(C-terminus) could be a common structural motif of membrane-disruptive peptides.