Publications
Our teams aspire to make discoveries that impact everyone, and core to our approach is sharing our research and tools to fuel progress in the field.
Our teams aspire to make discoveries that impact everyone, and core to our approach is sharing our research and tools to fuel progress in the field.
Sort By
1 - 15 of 10822 publications
mmMUSE: An mmWave-based Motion-resilient Universal Speech Enhancement System
Chenming He
Yanyong Zhang
Kai Wang
Dequan Wang
Lingyu Wang
the Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT), ACM (2026) (to appear)
Preview abstract
Voice-based smart systems can greatly enhance user experiences by allowing higher-quality interactions through better voice perception. Speech enhancement can benefit such systems by isolating noise from speech. Recently, integrating millimeter-wave (mmWave) with audio for speech perception has gained increasing attention due to microphones' limitations in noisy environments. However, mmWave-based vocal extraction is severely affected by motion, which disperses vocal signals across ranges and introduces distortions. In this paper, we propose an mmWave-based motion-resilient universal speech enhancement system called mmMUSE, which fuses mmWave and audio signals. To mitigate motion interference, we develop a Doppler-based method for motion-robust vocal signal extraction. Moreover, by introducing the Vocal-Noise-Ratio metric to assess the prominence of vocal signals from mmWave, we achieve real-time voice activity detection that gains 3.81 dB of SISDR in noisy speeches. Additionally, we design a two-stage complex-valued network that includes an attention-based fusion network for cross-modal complementing and a time-frequency masking network for correcting amplitude and phase of speech to isolate noises.
Using mmWave and audio datasets from 46 participants, mmMUSE outperforms the state-of-the-art speech enhancement models, achieving an average SISDR improvement of 3.12 dB. Additionally, mmMUSE achieves SISDR improvements of 16.51 dB, 17.93 dB, 14.93 dB, and 18.95 dB in controlled environments involving intense noise, extensive motion, multiple speakers, and various obstructive materials, respectively. Finally, we evaluate mmMUSE in real-world scenarios including running, public spaces, and driving, maintaining a word error rate (WER) below 10%.
View details
Preview abstract
For many practical applications of quantum computing, the slowest and most costly steps involve coherently accessing classical data. We help address this challenge by applying mass production techniques, which can sometimes allow us to perform operations many times in parallel for a cost that is comparable to a single execution[1-3]. We combine existing mass-production results with modern approaches for loading classical data using ``quantum read-only memory.'' We show that quantum mass production techniques offer no benefit when we consider a cost model that focuses purely on the number of non-Clifford gates. However, analyzing the constant factors in a more nuanced cost model, we find that it may be possible to obtain a reduction in cost of an order or magnitude or more for a variety reasonably-sized fault-tolerant quantum algorithms. We present several applications of quantum mass-production techniques beyond naive parallelization, including a strategy for reducing the cost of serial calls to the same data loading step.
View details
Preview abstract
AI coding assistants are rapidly becoming integral to modern software development. A key challenge in this space is the continual need to migrate and modernize codebases in response to evolving software ecosystems. Traditionally, such migrations have relied on rule-based systems and human intervention. With the advent of powerful large language models (LLMs), AI-driven agentic frameworks offer a promising alternative—but their effectiveness remains underexplored. In this paper, we introduce FreshBrew, a novel benchmark for evaluating AI-based agentic frameworks on project-level Java migrations. We benchmark several such frameworks, powered by state-of-the-art LLMs, and compare their performance against established rule-based tools. Our evaluation of AI agents on this benchmark of 228 repositories shows that the top-performing model, Gemini 2.5 Flash, can successfully migrate 56.5% of projects to JDK 17. Our empirical analysis reveals novel insights into the critical strengths and limitations of current agentic approaches, offering actionable insights into their real-world applicability. By releasing FreshBrew publicly upon acceptance, we aim to facilitate rigorous, reproducible evaluation and catalyze progress in AI-driven codebase modernization.
View details
Visualizing Dynamics of Charges and Strings in (2+1)D Lattice Gauge Theories
Tyler Cochran
Bernhard Jobst
Yuri Lensky
Gaurav Gyawali
Norhan Eassa
Melissa Will
Aaron Szasz
Dmitry Abanin
Rajeev Acharya
Laleh Beni
Trond Andersen
Markus Ansmann
Frank Arute
Kunal Arya
Abe Asfaw
Juan Atalaya
Brian Ballard
Alexandre Bourassa
Michael Broughton
David Browne
Brett Buchea
Bob Buckley
Tim Burger
Nicholas Bushnell
Anthony Cabrera
Juan Campero
Hung-Shen Chang
Jimmy Chen
Benjamin Chiaro
Jahan Claes
Agnetta Cleland
Josh Cogan
Roberto Collins
Paul Conner
William Courtney
Alex Crook
Ben Curtin
Sayan Das
Laura De Lorenzo
Paul Donohoe
ILYA Drozdov
Andrew Dunsworth
Alec Eickbusch
Aviv Elbag
Mahmoud Elzouka
Vinicius Ferreira
Ebrahim Forati
Austin Fowler
Brooks Foxen
Suhas Ganjam
Robert Gasca
Élie Genois
William Giang
Dar Gilboa
Raja Gosula
Alejo Grajales Dau
Dietrich Graumann
Alex Greene
Steve Habegger
Monica Hansen
Sean Harrington
Paula Heu
Oscar Higgott
Jeremy Hilton
Robert Huang
Ashley Huff
Bill Huggins
Cody Jones
Chaitali Joshi
Pavol Juhas
Hui Kang
Amir Karamlou
Kostyantyn Kechedzhi
Trupti Khaire
Bryce Kobrin
Alexander Korotkov
Fedor Kostritsa
John Mark Kreikebaum
Vlad Kurilovich
Dave Landhuis
Tiano Lange-Dei
Brandon Langley
Kim Ming Lau
Justin Ledford
Kenny Lee
Loick Le Guevel
Wing Li
Alexander Lill
Will Livingston
Aditya Locharla
Daniel Lundahl
Aaron Lunt
Sid Madhuk
Ashley Maloney
Salvatore Mandra
Leigh Martin
Orion Martin
Cameron Maxfield
Seneca Meeks
Anthony Megrant
Reza Molavi
Sebastian Molina
Shirin Montazeri
Ramis Movassagh
Charles Neill
Michael Newman
Murray Ich Nguyen
Chia Ni
Kris Ottosson
Alex Pizzuto
Rebecca Potter
Orion Pritchard
Ganesh Ramachandran
Matt Reagor
David Rhodes
Gabrielle Roberts
Kannan Sankaragomathi
Henry Schurkus
Mike Shearn
Aaron Shorter
Vladimir Shvarts
Vlad Sivak
Spencer Small
Clarke Smith
Sofia Springer
George Sterling
Jordan Suchard
Alex Sztein
Doug Thor
Mert Torunbalci
Abeer Vaishnav
Justin Vargas
Sergey Vdovichev
Guifre Vidal
Steven Waltman
Shannon Wang
Brayden Ware
Kristi Wong
Cheng Xing
Jamie Yao
Ping Yeh
Bicheng Ying
Juhwan Yoo
Grayson Young
Yaxing Zhang
Ningfeng Zhu
Yu Chen
Vadim Smelyanskiy
Adam Gammon-Smith
Frank Pollmann
Michael Knap
Nature, 642 (2025), 315–320
Preview abstract
Lattice gauge theories (LGTs) can be used to understand a wide range of phenomena, from elementary particle scattering in high-energy physics to effective descriptions of many-body interactions in materials. Studying dynamical properties of emergent phases can be challenging, as it requires solving many-body problems that are generally beyond perturbative limits. Here we investigate the dynamics of local excitations in a LGT using a two-dimensional lattice of superconducting qubits. We first construct a simple variational circuit that prepares low-energy states that have a large overlap with the ground state; then we create charge excitations with local gates and simulate their quantum dynamics by means of a discretized time evolution. As the electric field coupling constant is increased, our measurements show signatures of transitioning from deconfined to confined dynamics. For confined excitations, the electric field induces a tension in the string connecting them. Our method allows us to experimentally image string dynamics in a (2+1)D LGT, from which we uncover two distinct regimes inside the confining phase: for weak confinement, the string fluctuates strongly in the transverse direction, whereas for strong confinement, transverse fluctuations are effectively frozen. We also demonstrate a resonance condition at which dynamical string breaking is facilitated. Our LGT implementation on a quantum processor presents a new set of techniques for investigating emergent excitations and string dynamics.
View details
Scalable Private Partition Selection via Adaptive Weighting
Justin Y. Chen
Forty-second International Conference on Machine Learning (2025)
Preview abstract
In the differentially private partition selection problem (a.k.a. private set union, private key discovery), users hold subsets of items from an unbounded universe. The goal is to output as many items as possible from the union of the users' sets while maintaining user-level differential privacy. Solutions to this problem are a core building block for many privacy-preserving ML applications including vocabulary extraction in a private corpus, computing statistics over categorical data and learning embeddings over user-provided items.
We propose an algorithm for this problem, MaxAdaptiveDegree(MAD), which adaptively reroutes weight from items with weight far above the threshold needed for privacy to items with smaller weight, thereby increasing the probability that less frequent items are output. Our algorithm can be efficiently implemented in massively parallel computation systems allowing scalability to very large datasets. We prove that our algorithm stochastically dominates the standard parallel algorithm for this problem. We also develop a two-round version of our algorithm, MAD2R, where results of the computation in the first round are used to bias the weighting in the second round to maximize the number of items output. In experiments, our algorithms provide the best results across the board among parallel algorithms and scale to datasets with hundreds of billions of items, up to three orders of magnitude larger than those analyzed by prior sequential algorithms.
View details
The Cost of Consistency: Submodular Maximization with Constant Recourse
Paul Duetting
Federico Fusco
Ashkan Norouzi Fard
Ola Svensson
Proceedings of the 57th Annual ACM Symposium on Theory of Computing (2025), 1406–1417
Preview abstract
In this work, we study online submodular maximization and how the requirement of maintaining a stable solution impacts the approximation. In particular, we seek bounds on the best-possible
approximation ratio that is attainable when the algorithm is allowed to make, at most, a constant number of updates per step. We show a tight information-theoretic bound of $2/3$ for general monotone submodular functions and an improved (also tight) bound of $3/4$ for coverage functions. Since both these bounds are attained by non poly-time algorithms, we also give a poly-time randomized algorithm that achieves a $0.51$-approximation. Combined with an
information-theoretic hardness of $1/2$ for deterministic algorithms from prior work, our work thus shows a separation between deterministic and randomized algorithms, both information theoretically and for poly-time algorithms.
View details
Reasoning-SQL: Reinforcement Learning with Partial Rewards for Reasoning-Enhanced Text-to-SQL
Mohammadreza Pourreza
Shayan Talaei
Hailong Li
Azalia Mirhoseini
Amin Saberi
Conference on Language Modeling (COLM) (2025) (to appear)
Preview abstract
Text-to-SQL is a challenging task involving multiple reasoning-intensive subtasks, including natural language understanding, database schema comprehension, and precise SQL query formulation. Existing approaches often rely on handcrafted reasoning paths with inductive biases that can limit their overall effectiveness. Motivated by the recent success of reasoning-enhanced models such as DeepSeek R1 and OpenAI o1, which effectively leverage reward-driven self-exploration to enhance reasoning capabilities and generalization, we propose a novel set of partial rewards tailored specifically for the Text-to-SQL task. Our reward set includes schema-linking, AI feedback, n-gram similarity, and syntax check, explicitly designed to address the reward sparsity issue prevalent in reinforcement learning (RL). Leveraging group relative policy optimization (GRPO), our approach explicitly encourages large language models (LLMs) to develop intrinsic reasoning skills necessary for accurate SQL query generation. With models of different sizes, we demonstrate that RL-only training with our proposed rewards consistently achieves higher accuracy and superior generalization compared to supervised fine-tuning (SFT). Remarkably, our RL-trained 14B-parameter model significantly outperforms larger proprietary models, e.g. o3-mini by 4% and Gemini-1.5-Pro-002 by 3% on the BIRD benchmark. These highlight the efficacy of our proposed RL-training framework with partial rewards for enhancing both accuracy and reasoning capabilities in Text-to-SQL tasks.
View details
Applying multimodal AI to physiological waveforms improves genetic prediction of cardiovascular traits
Yuchen Zhou
Mahantesh I. Biradar
Jacqueline Shreibati
Dongbing Lai
Tae-Hwi Schwantes-An
Robert Luben
Zachary R. McCaw
Jorgen Engmann
Rui Providencia
Amand Floriaan Schmidt
Patricia B. Munroe
Howard Yang
Andrew Carroll
Anthony Khawaja
Babak Behsaz
American Journal of Human Genetics, 112 (2025), pp. 1562 - 1579
Preview abstract
Electronic health records, biobanks, and wearable biosensors enable the collection of multiple health modalities from many individuals. Access to multimodal health data provides a unique opportunity for genetic studies of complex traits because different modalities relevant to a single physiological system (e.g., circulatory system) encode complementary and overlapping information. We propose a multimodal deep learning method, multimodal representation learning for genetic discovery on low-dimensional embeddings (M-REGLE), for discovering genetic associations from a joint representation of complementary electrophysiological waveform modalities. M-REGLE jointly learns a lower representation (i.e., latent factors) of multimodal physiological waveforms using a convolutional variational autoencoder, performs genome-wide association studies (GWASs) on each latent factor, then combines the results to study the genetics of the underlying system. To validate the advantages of M-REGLE and multimodal learning, we apply it to common cardiovascular modalities (photoplethysmogram [PPG] and electrocardiogram [ECG]) and compare its results to unimodal learning methods in which representations are learned from each data modality separately but are statistically combined for downstream genetic comparison. M-REGLE identifies 19.3% more loci on the 12-lead ECG dataset, 13.0% more loci on the ECG lead I + PPG dataset, and its genetic risk score significantly outperforms the unimodal risk score at predicting cardiac phenotypes, such as atrial fibrillation (Afib), in multiple biobanks.
View details
Preview abstract
Despite exceptional achievements, training neural networks remains computationally expensive and is often plagued by instabilities that can degrade convergence. While learning rate schedules can help mitigate these issues, finding optimal schedules is time-consuming and resource-intensive. This work explores theoretical issues concerning training stability in the constant-learning-rate (i.e., without schedule) and small-batch-size regime. Surprisingly, we show that the order of gradient updates affects stability and convergence in gradient-based optimizers. We illustrate this new line of thinking using backward-SGD, which processes batch gradient updates like SGD but in reverse order. Our theoretical analysis shows that in contractive regions (e.g., around minima) backward-SGD converges to a point while the standard forward-SGD generally only converges to a distribution. This leads to improved stability and convergence which we demonstrate experimentally. While full backward-SGD is computationally intensive in practice, it highlights opportunities to exploit reverse training dynamics (or more generally alternate iteration orders) to improve training. To our knowledge, this represents a new and unexplored avenue in deep learning optimization.
View details
Preview abstract
Large-scale video generative models, capable of creating realistic videos of diverse visual concepts, are strong candidates for general-purpose physical world simulators. However, their adherence to physical commonsense across real-world actions remains unclear (e.g., playing tennis, backflip). Existing benchmarks suffer from limitations such as limited size, lack of human evaluation, sim-to-real gaps, and absence of fine-grained physical rule analysis. To address this, we introduce VideoPhy-2, an action-centric dataset for evaluating physical commonsense in generated videos. We curate 200 diverse actions and detailed prompts for video synthesis from modern generative models. We perform human evaluation that assesses semantic adherence, physical commonsense, and grounding of physical rules in the generated videos. Our findings reveal major shortcomings, with even the best model achieving only 22% joint performance (i.e., high semantic and physical commonsense adherence) on the hard subset of VideoPhy-2. We find that the models particularly struggle with conservation laws like mass and momentum. Finally, we also train VideoPhy-AutoEval, an automatic evaluator for fast, reliable assessment on our dataset. Overall, VideoPhy-2 serves as a rigorous benchmark, exposing critical gaps in video generative models and guiding future research in physically-grounded video generation. The data and code is available at https://videophy2.github.io/
View details
Scalability of Generative AI Models: Challenges and Opportunities in Large-Scale Data Generation and Training
International Journal of Computer Science and Information Technology Research (IJCSITR) (2025)
Preview abstract
Scalability of Generative AI Models: Challenges and Opportunities in Large-Scale Data Generation and Training
View details
I know what I don't know: improving model cascades through confidence tuning
Stephan Rabanser
Nathalie Rauschmayr
Petra Poklukar
Congchao Wang
2025
Preview abstract
Large-scale machine learning models deliver strong performance across a wide range of tasks but come with significant computational and resource constraints. To mitigate these challenges, local smaller models are often deployed alongside larger models, relying on routing and deferral mechanisms to offload complex tasks. However, existing approaches inadequately balance the capabilities of these models, often resulting in unnecessary deferrals or sub-optimal resource usage. In this work we introduce a novel loss function called Gatekeeper for calibrating smaller models in cascade setups. Our approach fine-tunes the smaller model to confidently handle tasks it can perform correctly while deferring complex tasks to the larger model. Moreover, it incorporates a mechanism for managing the trade-off between model performance and deferral accuracy, and is broadly applicable across various tasks and domains without any architectural changes. We evaluated our method on encoder-only, decoder-only, and encoder-decoder architectures. Experiments across image classification, language modeling, and vision-language tasks show that our approach substantially improves deferral performance.
View details
Preview abstract
We pioneer the study of in-context training for time-series foundation models. We create finetuning examples that not only include the usual (context, horizon) pairs for forecasting; but also related time-series examples in-context. We finetune a pretrained time-series foundation model on the type of in-context examples mentioned above. Our training is decoder-only and can adapt not only to any context, horizon pair (up to a certain maximum context) but also to any number of supplementary time-series examples (again up to a certain maximum number of examples). Appropriately trained models can then learn to borrow patterns from these related examples to do better on the original forecasting task. We show that this opens up interesting features like the ability to prompt the time-series foundation model with different related examples. This can help the finetuned model to adapt to specific features of a dataset at inference time. We show that such adaptions can lead to better zero-shot performance on popular forecasting benchmarks as compared to supervised deep learning methods, statistical models as well as other time-series foundation models.
View details
Multi-turn Function-calling via Graph-based Execution and Translation
Kai-Wei Chang
Ke Jiang
Jindong Gu
Fan Yin
2025
Preview abstract
We propose a principled method to synthesize high-quality multi-turn function calling trajectories to align large language model (LLM)-based agents. We start with iteratively building function calling graph and defining node operations to increase its complexity. This enables us to construct reliable reference. Then, based on the synthesized function calling graph, we adopt back-and-forth translation to first construct multi-turn user queries and then, fill in the function arguments with information in the query. We sample positive trajectories that distill the function graph reference and negative trajectories that contrast with the positive trajectories in targeted loss patterns in multi-turn scenarios. Training with the positive trajectories with supervised fine-tuning and preference optimization against negative trajectories, we obtain 67.42 on BFCL and 71.7 on ToolQuery with an open-sourced model with 14B parameters, surpassing the performance of strong proprietary models like o1.
View details
UWB Radar-based Heart Rate Monitoring: A Transfer Learning Approach
Elzbieta Gruzewska
Sebastien Baur
Matthew Baugh
Sharanya Srinivas
Matthew Thompson
Pramod Rudrapatna
Michael A. Sanchez
Lawrence Z. Cai
Timothy JA Chico
Robert F Storey
Emily Maz
Umesh Telang
Shravya Shetty
Mayank Daswani
arXiv (2025)
Preview abstract
Radar technology presents untapped potential for continuous, contactless, and passive heart rate monitoring via consumer electronics like mobile phones. However the variety of available radar systems and lack of standardization means that a large new paired dataset collection is required for each radar system. This study demonstrates transfer learning between frequency-modulated continuous wave (FMCW) and impulse-radio ultra-wideband (IR-UWB) radar systems, both increasingly integrated into consumer devices. FMCW radar utilizes a continuous chirp, while IR-UWB radar employs short pulses. Our mm-wave FMCW radar operated at 60 GHz with a 5.5 GHz bandwidth (2.7 cm resolution, 3 receiving antennas [Rx]), and our IR-UWB radar at 8 GHz with a 500 MHz bandwidth (30 cm resolution, 2 Rx). Using a novel 2D+1D ResNet architecture we achieved a mean absolute error (MAE) of 0.85 bpm and a mean absolute percentage error (MAPE) of 1.42% for heart rate monitoring with FMCW radar (N=119 participants, an average of 8 hours per participant). This model maintained performance (under 5 MAE/10% MAPE) across various body positions and heart rate ranges, with a 98.9% recall. We then fine-tuned a variant of this model, trained on single-antenna and single-range bin FMCW data, using a small (N=376, avg 6 minutes per participant) IR-UWB dataset. This transfer learning approach yielded a model with MAE 4.1 bpm and MAPE 6.3% (97.5% recall), a 25% MAE reduction over the IR-UWB baseline. This demonstration of transfer learning between radar systems for heart rate monitoring has the potential to accelerate its introduction into existing consumer devices.
View details
