Kevin Chan

Omnilingual ASR team, Gil Keren, Artyom Kozhevnikov et al.

Automatic speech recognition (ASR) has advanced in high-resource languages, but most of the world's 7,000+ languages remain unsupported, leaving thousands of long-tail languages behind. Expanding ASR coverage has been costly and limited by architectures that restrict language support, making extension inaccessible to most--all while entangled with ethical concerns when pursued without community collaboration. To transcend these limitations, we introduce Omnilingual ASR, the first large-scale ASR system designed for extensibility. Omnilingual ASR enables communities to introduce unserved languages with only a handful of data samples. It scales self-supervised pre-training to 7B parameters to learn robust speech representations and introduces an encoder-decoder architecture designed for zero-shot generalization, leveraging a LLM-inspired decoder. This capability is grounded in a massive and diverse training corpus; by combining breadth of coverage with linguistic variety, the model learns representations robust enough to adapt to unseen languages. Incorporating public resources with community-sourced recordings gathered through compensated local partnerships, Omnilingual ASR expands coverage to over 1,600 languages, the largest such effort to date--including over 500 never before served by ASR. Automatic evaluations show substantial gains over prior systems, especially in low-resource conditions, and strong generalization. We release Omnilingual ASR as a family of models, from 300M variants for low-power devices to 7B for maximum accuracy. We reflect on the ethical considerations shaping this design and conclude by discussing its societal impact. In particular, we highlight how open-sourcing models and tools can lower barriers for researchers and communities, inviting new forms of participation. Open-source artifacts are available at https://github.com/facebookresearch/omnilingual-asr.

Valentin Vie, Ryan Sheatsley, Sophia Beyda et al.

Planning algorithms are used in computational systems to direct autonomous behavior. In a canonical application, for example, planning for autonomous vehicles is used to automate the static or continuous planning towards performance, resource management, or functional goals (e.g., arriving at the destination, managing fuel fuel consumption). Existing planning algorithms assume non-adversarial settings; a least-cost plan is developed based on available environmental information (i.e., the input instance). Yet, it is unclear how such algorithms will perform in the face of adversaries attempting to thwart the planner. In this paper, we explore the security of planning algorithms used in cyber- and cyber-physical systems. We present two $\textit{adversarial planning}$ algorithms-one static and one adaptive-that perturb input planning instances to maximize cost (often substantially so). We evaluate the performance of the algorithms against two dominant planning algorithms used in commercial applications (D* Lite and Fast Downward) and show both are vulnerable to extremely limited adversarial action. Here, experiments show that an adversary is able to increase plan costs in 66.9% of instances by only removing a single action from the actions space (D* Lite) and render 70% of instances from an international planning competition unsolvable by removing only three actions (Fast Forward). Finally, we show that finding an optimal perturbation in any search-based planning system is NP-hard.

Zhongyuan Zhao, Yujun Ming, Kevin Chan et al.

Many routing and flow optimization problems in wired networks can be solved efficiently using minimum cost flow formulations. However, this approach does not extend to wireless multi-hop networks, where the assumptions of fixed link capacity and linear cost structure collapse due to contention for shared spectrum resources. The key challenge is that the long-term capacity of a wireless link becomes a non-linear function of its network context, including network topology, link quality, and the traffic assigned to neighboring links. In this work, we pursue a new direction of modeling wireless network under randomized medium access control by developing an analytical network digital twin (NDT) that predicts link duty cycles from network context. We generalize randomized contention as finding a Maximal Independent Set (MIS) on the conflict graph using weighted Luby's algorithm, derive an analytical model of link duty cycles, and introduce an iterative procedure that resolves the circular dependency among duty cycle, link capacity, and contention probability. Our numerical experiments show that the proposed NDT accurately predicts link duty cycles and congestion patterns with up to a 5000x speedup over packet-level simulation, and enables us to optimize link scheduling using gradient descent for reduced congestion and radio footprint.

Seohyeon Cha, Huancheng Chen, Dongjun Kim et al.

Large language models (LLMs) deliver robust performance across diverse applications, yet their deployment often faces challenges due to the memory and latency costs of storing and accessing billions of parameters. Post-training quantization (PTQ) enables efficient inference by mapping pretrained weights to low-bit formats without retraining, but its effectiveness depends critically on both the quantization objective and the rounding procedure used to obtain low-bit weight representations. In this work, we show that interpolating between symmetric and asymmetric calibration acts as a form of regularization that preserves the standard quadratic structure used in PTQ while providing robustness to activation mismatch. Building on this perspective, we derive a simple successive rounding procedure that naturally incorporates asymmetric calibration, as well as a bounded-search extension that allows for an explicit trade-off between quantization quality and the compute cost. Experiments across multiple LLM families, quantization bit-widths, and benchmarks demonstrate that the proposed bounded search based on a regularized asymmetric calibration objective consistently improves perplexity and accuracy over PTQ baselines, while incurring only modest and controllable additional computational cost.

Moin Aminnaseri, Farima Fatahi Bayat, Nikita Bhutani et al.

NL2SQL systems aim to address the growing need for natural language interaction with data. However, real-world information rarely maps to a single SQL query because (1) users express queries iteratively (2) questions often span multiple data sources beyond the closed-world assumption of a single database, and (3) queries frequently rely on commonsense or external knowledge. Consequently, satisfying realistic data needs require integrating heterogeneous sources, modalities, and contextual data. In this paper, we present Blue's Data Intelligence Layer (DIL) designed to support multi-source, multi-modal, and data-centric applications. Blue is a compound AI system that orchestrates agents and data for enterprise settings. DIL serves as the data intelligence layer for agentic data processing, to bridge the semantic gap between user intent and available information by unifying structured enterprise data, world knowledge accessible through LLMs, and personal context obtained through interaction. At the core of DIL is a data registry that stores metadata for diverse data sources and modalities to enable both native and natural language queries. DIL treats LLMs, the Web, and the User as source 'databases', each with their own query interface, elevating them to first-class data sources. DIL relies on data planners to transform user queries into executable query plans. These plans are declarative abstractions that unify relational operators with other operators spanning multiple modalities. DIL planners support decomposition of complex requests into subqueries, retrieval from diverse sources, and finally reasoning and integration to produce final results. We demonstrate DIL through two interactive scenarios in which user queries dynamically trigger multi-source retrieval, cross-modal reasoning, and result synthesis, illustrating how compound AI systems can move beyond single database NL2SQL.

Negar Erfaniantaghvayi, Zhongyuan Zhao, Kevin Chan et al.

Backpressure (BP) routing and its shortest-path biased variant (SP-BP) provide powerful congestion-aware multipath resource allocation for wireless multi-hop networks, but they rely on per-commodity queueing and slot-by-slot control that may be difficult to realize under practical or legacy forwarding architectures. Moreover, even state-of-the-art SP-BP still suffers from the last-packet problem when short-lived traffic coexists with streaming flows. To address these limitations, we propose Ant Backpressure (Ant-BP), a periodic and fully distributed routing scheme that decouples route learning from packet forwarding. Ant-BP uses virtual SP-BP to construct pheromone-based forwarding probabilities, while actual packets are forwarded through per-neighbor first-in-first-out (FIFO) queues with probabilistic next-hop selection. This architecture enables link-capacity sharing across commodities, mitigates starvation of short-lived traffic, and extends the benefits of SP-BP to network architectures based on per-neighbor FIFO forwarding. Through periodic virtual updates, Ant-BP also adapts to transient link failures and mobility-induced topology changes. Our theoretical analysis and simulations show that, compared with conventional ant colony optimization (ACO) routing, virtual SP-BP enables Ant-BP to establish higher-quality forwarding policies with lower overhead. As a result, Ant-BP improves latency and delivery ratio over SP-BP and ACO-based baselines under mixed streaming and bursty traffic, achieves throughput comparable to SP-BP at low and medium traffic load, and remains robust to mismatched virtual-traffic assumptions, transient link failures, and node mobility.

Rostyslav Olshevskyi, Zhongyuan Zhao, Kevin Chan et al.

Graph neural networks (GNNs) are powerful tools for developing scalable, decentralized artificial intelligence in large-scale networked systems, such as wireless networks, power grids, and transportation networks. Currently, GNNs in networked systems mostly follow a paradigm of `centralized training, distributed execution', which limits their adaptability and slows down their development cycles. In this work, we fill this gap for the first time by developing a communication-efficient, fully distributed online training approach for GNNs applied to large networked systems. For a mini-batch with $B$ samples, our approach of training an $L$-layer GNN only adds $L$ rounds of message passing to the $LB$ rounds required by GNN inference, with doubled message sizes. Through numerical experiments in graph-based node regression, power allocation, and link scheduling in wireless networks, we demonstrate the effectiveness of our approach in training GNNs under supervised, unsupervised, and reinforcement learning paradigms.

Seohyeon Cha, Kevin Chan, Gustavo de Veciana et al.

The growing demand for intelligent services on resource-constrained edge devices has spurred the development of collaborative inference systems that distribute workloads across end devices, edge servers, and the cloud. While most existing frameworks focus on single-task, single-model scenarios, many real-world applications (e.g., autonomous driving and augmented reality) require concurrent execution of diverse tasks including detection, segmentation, and depth estimation. In this work, we propose a unified framework to jointly decide which multi-task models to deploy (onload) at clients and edge servers, and how to route queries across the hierarchy (offload) to maximize overall inference accuracy under memory, compute, and communication constraints. We formulate this as a mixed-integer program and introduce J3O (Joint Optimization of Onloading and Offloading), an alternating algorithm that (i) greedily selects models to onload via Lagrangian-relaxed submodular optimization and (ii) determines optimal offloading via constrained linear programming. We further extend J3O to account for batching at the edge, maintaining scalability under heterogeneous task loads. Experiments show J3O consistently achieves over $97\%$ of the optimal accuracy while incurring less than $15\%$ of the runtime required by the optimal solver across multi-task benchmarks.

Jake B. Perazzone, Shiqiang Wang, Mingyue Ji et al.

Federated learning (FL) is a useful tool that enables the training of machine learning models over distributed data without having to collect data centrally. When deploying FL in constrained wireless environments, however, intermittent connectivity of devices, heterogeneous connection quality, and non-i.i.d. data can severely slow convergence. In this paper, we consider FL with arbitrary device participation probabilities for each round and show that by weighing each device's update by the reciprocal of their per-round participation probability, we can guarantee convergence to a stationary point. Our bound applies to non-convex loss functions and non-i.i.d. datasets and recovers state-of-the-art convergence rates for both full and uniform partial participation, including linear speedup, with only a single-sided learning rate. Then, using the derived convergence bound, we develop a new online client selection and power allocation algorithm that utilizes the Lyapunov drift-plus-penalty framework to opportunistically minimize a function of the convergence bound and the average communication time under a transmit power constraint. We use optimization over manifold techniques to obtain a solution to the minimization problem. Thanks to the Lyapunov framework, one key feature of the algorithm is that knowledge of the channel distribution is not required and only the instantaneous channel state information needs to be known. Using the CIFAR-10 dataset with varying levels of data heterogeneity, we show through simulations that the communication time can be significantly decreased using our algorithm compared to uniformly random participation, especially for heterogeneous channel conditions.

Jake Perazzone, Shiqiang Wang, Mingyue Ji et al.

Federated learning (FL) is a useful tool in distributed machine learning that utilizes users' local datasets in a privacy-preserving manner. When deploying FL in a constrained wireless environment; however, training models in a time-efficient manner can be a challenging task due to intermittent connectivity of devices, heterogeneous connection quality, and non-i.i.d. data. In this paper, we provide a novel convergence analysis of non-convex loss functions using FL on both i.i.d. and non-i.i.d. datasets with arbitrary device selection probabilities for each round. Then, using the derived convergence bound, we use stochastic optimization to develop a new client selection and power allocation algorithm that minimizes a function of the convergence bound and the average communication time under a transmit power constraint. We find an analytical solution to the minimization problem. One key feature of the algorithm is that knowledge of the channel statistics is not required and only the instantaneous channel state information needs to be known. Using the FEMNIST and CIFAR-10 datasets, we show through simulations that the communication time can be significantly decreased using our algorithm, compared to uniformly random participation.

Todd Huster, Jeremy E. J. Cohen, Zinan Lin et al.

Generative adversarial networks (GANs) are often billed as "universal distribution learners", but precisely what distributions they can represent and learn is still an open question. Heavy-tailed distributions are prevalent in many different domains such as financial risk-assessment, physics, and epidemiology. We observe that existing GAN architectures do a poor job of matching the asymptotic behavior of heavy-tailed distributions, a problem that we show stems from their construction. Additionally, when faced with the infinite moments and large distances between outlier points that are characteristic of heavy-tailed distributions, common loss functions produce unstable or near-zero gradients. We address these problems with the Pareto GAN. A Pareto GAN leverages extreme value theory and the functional properties of neural networks to learn a distribution that matches the asymptotic behavior of the marginal distributions of the features. We identify issues with standard loss functions and propose the use of alternative metric spaces that enable stable and efficient learning. Finally, we evaluate our proposed approach on a variety of heavy-tailed datasets.

Benjamin A. Miller, Mustafa Çamurcu, Alexander J. Gomez et al.

Vertex classification is vulnerable to perturbations of both graph topology and vertex attributes, as shown in recent research. As in other machine learning domains, concerns about robustness to adversarial manipulation can prevent potential users from adopting proposed methods when the consequence of action is very high. This paper considers two topological characteristics of graphs and explores the way these features affect the amount the adversary must perturb the graph in order to be successful. We show that, if certain vertices are included in the training set, it is possible to substantially an adversary's required perturbation budget. On four citation datasets, we demonstrate that if the training set includes high degree vertices or vertices that ensure all unlabeled nodes have neighbors in the training set, we show that the adversary's budget often increases by a substantial factor---often a factor of 2 or more---over random training for the Nettack poisoning attack. Even for especially easy targets (those that are misclassified after just one or two perturbations), the degradation of performance is much slower, assigning much lower probabilities to the incorrect classes. In addition, we demonstrate that this robustness either persists when recently proposed defenses are applied, or is competitive with the resulting performance improvement for the defender.

Stephen Pasteris, Fabio Vitale, Kevin Chan et al.

We introduce a new online learning framework where, at each trial, the learner is required to select a subset of actions from a given known action set. Each action is associated with an energy value, a reward and a cost. The sum of the energies of the actions selected cannot exceed a given energy budget. The goal is to maximise the cumulative profit, where the profit obtained on a single trial is defined as the difference between the maximum reward among the selected actions and the sum of their costs. Action energy values and the budget are known and fixed. All rewards and costs associated with each action change over time and are revealed at each trial only after the learner's selection of actions. Our framework encompasses several online learning problems where the environment changes over time; and the solution trades-off between minimising the costs and maximising the maximum reward of the selected subset of actions, while being constrained to an action energy budget. The algorithm that we propose is efficient and general in that it may be specialised to multiple natural online combinatorial problems.

Shiqiang Wang, Tiffany Tuor, Theodoros Salonidis et al.

Emerging technologies and applications including Internet of Things (IoT), social networking, and crowd-sourcing generate large amounts of data at the network edge. Machine learning models are often built from the collected data, to enable the detection, classification, and prediction of future events. Due to bandwidth, storage, and privacy concerns, it is often impractical to send all the data to a centralized location. In this paper, we consider the problem of learning model parameters from data distributed across multiple edge nodes, without sending raw data to a centralized place. Our focus is on a generic class of machine learning models that are trained using gradient-descent based approaches. We analyze the convergence bound of distributed gradient descent from a theoretical point of view, based on which we propose a control algorithm that determines the best trade-off between local update and global parameter aggregation to minimize the loss function under a given resource budget. The performance of the proposed algorithm is evaluated via extensive experiments with real datasets, both on a networked prototype system and in a larger-scale simulated environment. The experimentation results show that our proposed approach performs near to the optimum with various machine learning models and different data distributions.

Zongqing Lu, Swati Rallapalli, Kevin Chan et al.

Convolutional Neural Networks (CNNs) have revolutionized the research in computer vision, due to their ability to capture complex patterns, resulting in high inference accuracies. However, the increasingly complex nature of these neural networks means that they are particularly suited for server computers with powerful GPUs. We envision that deep learning applications will be eventually and widely deployed on mobile devices, e.g., smartphones, self-driving cars, and drones. Therefore, in this paper, we aim to understand the resource requirements (time, memory) of CNNs on mobile devices. First, by deploying several popular CNNs on mobile CPUs and GPUs, we measure and analyze the performance and resource usage for every layer of the CNNs. Our findings point out the potential ways of optimizing the performance on mobile devices. Second, we model the resource requirements of the different CNN computations. Finally, based on the measurement, pro ling, and modeling, we build and evaluate our modeling tool, Augur, which takes a CNN configuration (descriptor) as the input and estimates the compute time and resource usage of the CNN, to give insights about whether and how e ciently a CNN can be run on a given mobile platform. In doing so Augur tackles several challenges: (i) how to overcome pro ling and measurement overhead; (ii) how to capture the variance in different mobile platforms with different processors, memory, and cache sizes; and (iii) how to account for the variance in the number, type and size of layers of the different CNN configurations.

Ashwin Bahulkar, Boleslaw K. Szymanski, Nitesh Chawla et al.

We study a unique network dataset including periodic surveys and electronic logs of dyadic contacts via smartphones. The participants were a sample of freshmen entering university in the Fall 2011. Their opinions on a variety of political and social issues and lists of activities on campus were regularly recorded at the beginning and end of each semester for the first three years of study. We identify a behavioral network defined by call and text data, and a cognitive network based on friendship nominations in ego-network surveys. Both networks are limited to study participants. Since a wide range of attributes on each node were collected in self-reports, we refer to these networks as attribute-rich networks. We study whether student preferences for certain attributes of friends can predict formation and dissolution of edges in both networks. We introduce a method for computing student preferences for different attributes which we use to predict link formation and dissolution. We then rank these attributes according to their importance for making predictions. We find that personal preferences, in particular political views, and preferences for common activities help predict link formation and dissolution in both the behavioral and cognitive networks.