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 11415 publications
Preview abstract
In a prior column, we wrote about how measuring productivity can be viewed as a form of modeling and that all models are wrong, but some are useful. That discussion centered on the idea of ensuring that a productivity model was inclusive of multiple metrics and that those metrics covered the various facets of productivity and covered each facet reasonably well. In that article, we set aside the question of what makes a good individual productivity metric that can be combined with others into a (hopefully) useful model of productivity. In this article, we’ll share some things we consider when building an individual metric, including an example of a novel metric we built in the aftermath of the COVID pandemic.
View details
Understanding U.S. Users' Security and Privacy Transparency Needs for Consumer-Facing Generative AI
Jiaxun Cao
Yu Dong
Chunxi Zhan
Rithvik Neti
Pardis Emami-Naeini
USENIX Symposium on Usable Privacy and Security (SOUPS) (2026)
Preview abstract
Users increasingly rely on consumer-facing generative AI (GenAI) for tasks ranging from everyday needs to sensitive use cases. Yet, it remains unclear whether and how existing security and privacy (S&P) communications in GenAI tools shape users’ adoption decisions and experiences. Understanding how users seek, interpret, and evaluate S&P information is critical for designing usable transparency that users can trust and act on. We conducted semi-structured interviews and design sessions with 21 U.S. GenAI users. Our findings suggest that available S&P information rarely drove initial adoption in practice, as participants often perceived it as incomplete, ineffective, or not credible. Instead, they relied on rough proxies (e.g., popularity) to infer S&P practices. After adoption, S&P uncertainty constrained participants’ willingness to use GenAI tools, especially for high-stakes purposes, and, in some cases, contributed to discontinued use. Participants therefore called for transparency that supports decisions and actions through trustworthy information (e.g., independent evaluations) and usable interfaces (e.g., on-demand disclosure). We categorize participants’ desired design practices into five dimensions to facilitate systematic future investigation into best practices. We conclude with recommendations for researchers, designers, and policymakers to improve S&P transparency in consumer-facing GenAI.
View details
Preview abstract
We study the stochastic linear bandits with parameter noise model, in which the reward of action $a$ is $a^\top \theta$ where $\theta$ is sampled i.i.d.
We show a regret upper bound of $\tO \brk{\sqrt{d T \logKdelta \sigmax}}$ for a horizon $T$, general action set of size $K$ of dimension $d$, and where $\sigmax$ is the maximal variance of the reward for any action.
We further provide a lower bound of $\tOmega \brk{d \sqrt{T \sigmax}}$ which is tight (up to logarithmic factors) whenever $\log \brk{K} \approx d$.
For more specific action sets, $\ell_p$ unit balls with $p \leq 2$ and dual norm $q$, we show that the minimax regret is $\tTheta \brk{\sqrt{dT \vsigq}}$, where $\vsigq$ is a variance-dependent quantity that is always at most $4$.
This is in contrast to the minimax regret attainable for such sets in the classic additive noise model, where the regret is of order $d \sqrt{T}$.
Surprisingly, we show that this optimal (up to logarithmic factors) regret bound is attainable using a very simple explore-exploit algorithm.
View details
Preview abstract
We introduce AASE (Activation-based AI Safety Enforcement), a framework for post-perception safety monitoring in large language models. Unlike pre-perception approaches that analyze input or output text, AASE monitors the model's internal activation patterns—what the model "understands" rather than what text it processes or generates—enabling detection of safety-relevant states before harmful outputs are produced. The framework comprises three techniques: Activation Fingerprinting (AF) for harmful content detection, Agent Action Gating (AAG) for prompt injection defense, and Activation Policy Compliance (APC) for enterprise policy enforcement. We introduce paired contrastive training to isolate safety-relevant signals from confounding factors such as topic and style, addressing signal entanglement in polysemantic activations. Validation across 7 models from 3 architecture families shows strong class separation: Gemma-2-9B achieves AUC 1.00 with 7.2σ separation across all probes; AAG achieves AUC ≥0.88 across all models on the InjecAgent benchmark; APC achieves 0.97-1.00 AUC across three enterprise policies. Model size correlates with probe quality—Gemma-2-9B (7.2σ separation) outperforms Gemma-2-2B (4.3σ). All techniques survive INT4 quantization with minimal separation degradation. AASE is 9× faster than Llama Guard 3 (33ms vs 306ms) with higher TPR (88% vs 50%) at a tunable threshold that trades FPR for detection sensitivity, adding only 0.002ms probe overhead to existing inference.
View details
Preview abstract
Modern user interfaces are complex composites, with elements originating from various sources, such as the operating system, apps, a web browser, or websites. Many security and privacy models implicitly depend on users correctly identifying an element's source, a concept we term ''surface attribution.'' Through two large-scale vignette-based surveys (N=4,400 and N=3,057), we present the first empirical measurement of this ability.
We find that users struggle, correctly attributing UI source only 55% of the time on desktop and 53% on mobile. Familiarity and strong brand cues significantly improve accuracy, whereas UI positioning, a long-held security design concept especially for browsers, has minimal impact. Furthermore, simply adding a ''Security & Privacy'' brand cue to Android permission prompts failed to improve attribution. These findings demonstrate a fundamental gap in users' mental models, indicating that relying on them to distinguish trusted UI is a fragile security paradigm.
View details
Impact of User Tonalities on Coding AI Model Performance and Behavior
Selva Jagannathan
Alec Mirchandani
Discover Artificial Intelligence (2026)
Preview abstract
This study investigates how different user tones—rude, polite, threatening, sarcastic, and neutral—affect the performance and behavior of coding AI models like Claude Opus 4.6, Qwen 3.5, and CodeGemma. We tracked technical metrics like correctness and adherence alongside behavioral traits such as sycophancy and defensiveness. Our findings show that tone has a direct impact on model reliability, with neutral prompts achieving the peak average correctness score (5.88 on a 1–10 scale). In contrast, non-neutral stimuli incurred a measurable "correctness tax": sarcasm led to a 9.0% drop in correctness and an 82.5% jump in sycophancy, while polite prompts caused a 4.7% decrease in correctness alongside a 36.4% increase in sycophancy. Threatening tones resulted in a 1.5% loss in correctness and made models 52.9% more defensive. These results suggest that user tonality is a key factor linking technical accuracy to behavioral stability. We recommend using professional, neutral prompting to ensure better logic and less bias in AI-assisted coding, and we propose "Tonal Resilience" as a necessary benchmark for future model evaluations.
View details
A Theoretical Framework for Modular Learning of Robust Generative Models
Proceedings of the 43rd International Conference on Machine Learning (ICML 2026)
Preview abstract
Training large-scale generative models is resource-intensive and relies heavily on heuristic dataset weighting. We address two fundamental questions: Can we train Large Language Models (LLMs) modularly, combining small, domain-specific experts to match monolithic performance, and can we do so robustly for any data mixture, eliminating heuristic tuning? We present a theoretical framework for modular generative modeling where a set of pre-trained experts are combined via a gating mechanism. We define the space of normalized gating functions $\mathcal{G}_{1}$ and formulate the problem as a minimax game to find a single robust gate that minimizes divergence to the worst-case data mixture. We prove the existence of such a robust gate using Kakutani's fixed-point theorem and show that modularity acts as a strong regularizer, with generalization bounds scaling with the lightweight gate's complexity. Furthermore, we prove that this modular approach can theoretically outperform models retrained on aggregate data, with the gap characterized by the Jensen-Shannon Divergence. Finally, we introduce a scalable Stochastic Primal-Dual algorithm and a Structural Distillation method for efficient inference. Empirical results on synthetic and real-world datasets confirm that our modular architecture effectively mitigates gradient conflict and can robustly outperform monolithic baselines.
View details
Agentic AI Infrastructure in Practice: Learn These Key Hurdles to Deploy Production AI Agents Efficiently
https://swisscognitive.ch/ (2026)
Preview abstract
The emergence of Agentic AI—autonomous systems capable of reasoning, decision-making, and multi-step execution—represents a paradigm shift in enterprise technology. Moving beyond simple generative tasks, these agents offer the potential to solve long-standing industry pain points, with over 90% of enterprises planning integration within the next three years. However, the transition from successful proof-of-concept (PoC) to a resilient, production-grade system presents significant hurdles.
This article categorizes these challenges into three primary domains:
Technical and Engineering Hurdles: Issues such as "entangled workflows" that complicate debugging, the struggle to maintain output quality and mitigate hallucinations, and the unpredictability caused by shifting underlying models or data sources.
People, Process, and Ecosystem Hurdles: The high operational costs and unclear ROI of large models, the necessity of a new "Agent Ops" skillset, the complexity of integrating agents with disparate enterprise systems, and a rapidly evolving regulatory landscape.
The Pace of Change and Security risks: The technical debt incurred by shifting software frameworks and the expanded attack surface created by autonomous agents.
The article concludes that successful deployment requires a shift from informal "vibe-testing" to rigorous engineering discipline. By adopting code-first frameworks, establishing robust evaluation metrics (KPIs), and prioritizing functional deployment over theoretical optimization, organizations can effectively manage the lifecycle of Agentic AI and realize its transformative business value.
View details
What does your wearable know about the festive season?
Justin Phillips
Abram Schönfeldt
YongSuk Cho
Conor Heneghan
Robert Harle
(2026)
Preview abstract
As we reach the end of the year and people look forward to spending quality time with loved ones, here at Fitbit, we wonder what our Pixel watches and Fitbit trackers can tell us about how we are spending the festive season. We looked at the data of 11.8 million of our users all over the world between January 2022 and July 2025. Here are the key stats we wanted to share with you!
View details
Preview abstract
The rapid expansion of the Internet of Things (IoT) and smart home ecosystems has led to a fragmented landscape of user data management across consumer electronics (CE) such as Smart TVs, gaming consoles, and set-top boxes. Current onboarding processes on these devices are characterized by high friction due to manual data entry and opaque data-sharing practices. This paper introduces the User Data Sharing System (UDSS), a platform-agnostic framework designed to facilitate secure, privacy-first PII (Personally Identifiable Information) exchange between device platforms and third-party applications. Our system implements a Contextual Scope Enforcement (CSE) mechanism that programmatically restricts data exposure based on user intent—specifically distinguishing between Sign-In and Sign-Up workflows. Unlike cloud-anchored identity standards such as FIDO2/WebAuthn, UDSS is designed for shared, device-centric CE environments where persistent user-to-device bind-ing cannot be assumed. We further propose a tiered access model that balances developer needs with regulatory compliance (GDPR/CCPA). A proof-of-concept implementation on a reference ARMv8 Linux-based middleware demonstrates that UDSS reduces user onboarding latency by 65% and measurably reduces PII over-exposure risk through protocol-enforced data minimization. This framework provides a standardized approach to identity management in the heterogeneous CE market.
View details
Preview abstract
We address the problem of training conversion prediction models in advertising domains under privacy constraints, where direct links between ad clicks and conversions are unavailable. Motivated by privacy-preserving browser APIs and the deprecation of third-party cookies, we study a setting where the learner observes a sequence of clicks and a sequence of conversions, but can only link a conversion to a set of candidate clicks (an attribution set) rather than a unique source. We formalize this as learning from attribution sets generated by an oblivious adversary equipped with a prior distribution over the candidates. Despite the lack of explicit labels, we construct an unbiased estimator of the population loss from these coarse signals via a novel approach. Leveraging this estimator, we show that Empirical Risk Minimization achieves generalization guarantees that scale with the informativeness of the prior and is also robust against estimation errors in the prior, despite complex dependencies among attribution sets.
Simple empirical evaluations on standard datasets suggest our unbiased approach significantly outperforms common industry heuristics, particularly in regimes where attribution sets are large or overlapping.
View details
Pixel Watch: Robust Heart Rate Sensing from Multipath PPG and On-Device Deep Learning Trained on 10,000 hours of Free-Living and Fitness Data
Megan Walker
Yojan Patel
Shyam Tailor
Matt Wimmer
Brennan Garrett
Dan Howe
Hamed Vavadi
Tien Le
Steve Diamond
Oleksiy Vyalov
Vik Sharma
Pete Richards
Tracy Giest
Erika Siegel
Tuan Phan
Sam Mravca
Derrick Vickers
Benjamin Stone
Justin Phillips
YongSuk Cho
Stefanie Hollidge
Antony Siahaan
Soren Brage
Shwetak Patel
Robert Harle
IEEE Sensors Letters (2026)
Preview abstract
The Pixel Watch 2 (PW2) is the first Google smartwatch to combine multipath photoplethysmography (PPG) with deep learning-based heart rate inference, designed to significantly improve sensing accuracy during motion-heavy activities. The device processes 10 optical channels using an on-device, 15-layer temporally dilated convolutional neural network (~300K parameters) to yield a 1 Hz heart rate output. Crucial to this model's performance was its training on a massive dataset comprising 10,000 hours of data from 962 participants, curated from a broader corpus of controlled and free-living activities. We evaluated the PW2's sensing performance across two independent validation sets: an in-house fitness dataset (229 participants, 250 hours) and an external free-living dataset (27 participants, 1000+ hours). The system achieved 95% Limits of Agreement of -10.34 to 8.66 BPM during exercise and -6.57 to 7.48 BPM during free-living activities, demonstrating substantially tighter error margins than previous Google devices. Finally, we discuss key design lessons, emphasizing that large-scale deep learning was instrumental in fully leveraging multipath PPG hardware over traditional signal processing approaches.
View details
Preview abstract
We study the fair allocation of indivisible goods with variable groups. In this model, the goal is to partition the agents into groups of given sizes and allocate the goods to the groups in a fair manner. We show that for any number of groups and corresponding sizes, there always exists an envy-free up to one good (EF1) outcome, thereby generalizing an important result from the individual setting. Our result holds for arbitrary monotonic utilities and comes with an efficient algorithm. We also prove that the EF1 existence can be guaranteed even when the goods lie on a path and each group must receive a connected bundle. In addition, we consider a probabilistic model where the utilities are additive and drawn randomly from a distribution. We show that if there are n agents and the number of goods m is divisible by the number of groups k, then an envy-free outcome exists with high probability if m = ω(log n), and this bound is tight. On the other hand, if m is not divisible by k, then an envy-free outcome is unlikely to exist as long as m = o(√n).
View details
Multi-Agent Design: Optimizing Agents with Better Prompts and Topologies
Han Zhou
Shariq Iqbal
Ivan Vulić
Anna Korhonen
International Conference on Learning Representations (ICLR) (2026)
Preview abstract
Large language models (LLMs), employed as multiple agents that interact and collaborate with each other, have excelled at solving complex tasks. The agents are programmed with {prompts} that declare their functionality, along with the {workflows} that orchestrate interactions within a structured flow. Designing prompts and workflows for multi-agent systems is inherently complex, especially when addressing a new task. It often demands expert-level knowledge and involves significant trial and error. Gaining a deep understanding of the factors that contribute to effective multi-agent systems is essential for automating the entire process. Motivated by this, we first conduct an in-depth analysis of the design spaces for multi-agent systems, focusing on the impact of prompts, scaling the number of agents, and common types of agentic modules. Our findings reveal that top-performing systems often emerge from simpler design spaces, where prompts play a critical role in enhancing agent functionality and enabling more effective scaling. Based on the insights, we propose Multi-Agent System Search (MASS), a multi-stage optimization framework that performs the optimization in a pruned design space, with prompts and an influential subset of modules. We show that MASS-optimized multi-agent systems outperform existing alterntives by a substantial margin. Based on the MASS-found systems, we finally propose design principles behind building effective multi-agent systems.
View details
Leveraging LLMs to Understand Public Perception of Earthquake Early Warnings: A Case Study of the 2025 M6.2 Türkiye Earthquake
Hanjing Wang
Marc Stogaitis
Youngmin Cho
Richard Allen
Boone Spooner
Patrick Robertson
Greg Wimpey
Robert Bosch
Nivetha Thiruverahan
Steve Malkos
Alexei Barski
Tajinder Gadh
Scientific Reports, 16 (2026), pp. 22226
Preview abstract
Android's Earthquake Alert (AEA) system provided life-saving early warnings to millions during the M6.2 Marmara Ereğlisi, Türkiye earthquake on April 23, 2025. This event, the largest in the region in 25 years, served as a critical real-world test for smartphone-based Earthquake Early Warning (EEW) systems. The AEA system successfully delivered alerts to users with high precision, offering over a minute of warning before the strongest shaking reached urban areas. This study leveraged Large Language Models (LLMs) to analyze 560 public social media posts from the X platform, extracting 42 distinct attributes related to user experience and behavior. Statistical analyses revealed significant relationships, notably a strong correlation between user trust and alert timeliness. Key findings indicate that alerts received before shaking, clear information, and audible notifications significantly increased the likelihood of active protective responses and enhanced perceived usefulness and future trust in the system. Conversely, annoyance and perceived inaccuracy led to inaction and diminished trust. The study highlights the profound impact of smartphone-based EEW in mitigating seismic risk, providing actionable insights for optimizing alert design, public education campaigns, and future behavioral research to improve the effectiveness of such systems in seismically active regions.
View details
