arXiv subjects

Brian Wang

Publications and source records attributed to Brian Wang.

11 recordsLinked to original sources

arXiv2026

FrontierChallenge: Evaluating Scientific Workflow Completion

Scientific agents increasingly analyze data, execute code, and produce research artifacts, yet most benchmarks emphasize final answers, isolated programs, or a single domain. We introduce FrontierChallenge, a cross-domain benchmark comprising 300 end-to-end scientific workflows. In this paper, we release and evaluate 97 of these tasks, spanning quantum chemistry, molecular dynamics, materials characterization, analytical chemistry, life science, and electrochemistry/environment. Each task provides fixed inputs and specifies a bundle of required scientific deliverables. We evaluate twelve frontier models with three agent scaffolds. Pass Rate measures the fraction of tasks satisfying the full-completion criterion, while Avg. Score captures partial progress. Each of the best-performing configurations completed only 20 of the 97 released tasks, yielding a Pass Rate of 20.6%. Partial progress translated especially poorly into complete delivery in analytical chemistry and electrochemistry/environment: Avg. Scores reached 87.6 and 94.9, but the highest Pass Rates were only 4% and 0%. Among non-passing Claude Code trajectories, 75.5% still ended with language claiming completion. Complementary HDS6 process scores correlate strongly with task outcomes, supporting FrontierChallenge as a benchmark of Heavy Duty Solver capabilities. These findings show that neither high partial scores nor confident claims of completion reliably indicate that a scientific task has been fully delivered, highlighting the need to evaluate end-to-end workflow execution and the completeness of scientific deliverables together.

cs.AI↗

arXiv2026

Apodex Discovery: Reality Benchmarks and Environments for Evaluating and Building Discoverative Artificial Intelligence

Apollo did not reach the Moon merely because its engineers could solve difficult equations. It succeeded by turning a distant ambition into a mission architecture of explicit objectives, simulation, verification, and repeated correction. AI now faces a similar transition: frontier models can solve difficult tasks once the problem, tools, and success criteria are specified, yet consequential real-world challenges rarely arrive in an executable or verifiable form. We introduce Apodex Discovery, a framework for building and evaluating discoverative AI through the heavy-duty solver, a system comprising a foundation model, harness, tools, and control policies that pursues extended, stateful, verifiable investigations. It has three core components. First, a problem-scouting process surveyed 561 industries across 16 sectors, assembled 423 high-value real-world problems, and selected 20 for the initial release. Second, a common environment-task-episode abstraction provides data, tools, constraints, feedback, trajectory recording, and verification of intermediate artifacts and final submissions. Third, HDS6 evaluates Tools, Repair, Alternatives, Coherence, Evidence, and Scope independently of final-task success. In AAV capsid design, Apodex surpassed the published state of the art by 7% across viability, tropism, structure prediction, and generative design. In drug repurposing and reformulation, a task-specific biomedical environment improved the mean normalized prediction score of GPT-5.5 and GPT-5.6-sol by 2.5 and 7.6 points over the same closed-book backbone. Controlled ablations show that the fixed TRACES episode interface enables attribution of performance differences to specific solver components. Apodex Discovery moves AI evaluation beyond predefined benchmarks toward verifiable investigations aimed at genuine discovery.

cs.AI↗

arXiv2026

CADET: A Modular Platform for Evaluating Distributed Cooperative Autonomy in Connected Autonomous Vehicles

Deep learning models are increasingly central to autonomous vehicle (AV) pipelines, yet their integration has traditionally followed a monolithic design where perception, planning, and control execute on a single onboard computer. This design overlooks the emerging paradigm of cooperative autonomy, where vehicles interact with roadside units (RSUs), edge servers, and cloud-hosted intelligence through vehicle-to-everything (V2X) connectivity. Cooperative perception and control improve safety and efficiency, but also introduce systems-level challenges: network latency, compute heterogeneity, and multi-tenant contention, all critically affect real-time decision-making. These challenges are further amplified by the increasing reliance on large foundation models, whose scale necessitates cloud deployment. We present CADET (Cooperative Autonomy through Distributed Experimentation Toolkit), a modular platform for systematic and reproducible evaluation of distributed cooperative autonomy systems under realistic deployment conditions. CADET decouples the AV stack into composable modules that can be flexibly deployed across vehicles, infrastructure, and edge/cloud tiers. The framework integrates state-of-the-art models, incorporates trace-driven network and workload emulation, and provides synchronized model-, system-, and task-level instrumentation. Through V2V and V2I experiments, we show that distributed deployment choices fundamentally shape safety, with V2V intent packets outperforming cloud-based perception and RSU-assisted perception sustaining safety until overloaded by concurrent requests. Although designed for AV pipelines, CADET also supports dataset-driven experimentation, enabling systems and ML researchers to benchmark distributed inference workloads independently of full vehicle simulation. CADET is open source, with code and demo available at https://nesl.github.io/cadet-web.

cs.RO↗

arXiv2025

SIGMUS: Semantic Integration for Knowledge Graphs in Multimodal Urban Spaces

Modern urban spaces are equipped with an increasingly diverse set of sensors, all producing an abundance of multimodal data. Such multimodal data can be used to identify and reason about important incidents occurring in urban landscapes, such as major emergencies, cultural and social events, as well as natural disasters. However, such data may be fragmented over several sources and difficult to integrate due to the reliance on human-driven reasoning for identifying relationships between the multimodal data corresponding to an incident, as well as understanding the different components which define an incident. Such relationships and components are critical to identifying the causes of such incidents, as well as producing forecasting the scale and intensity of future incidents as they begin to develop. In this work, we create SIGMUS, a system for Semantic Integration for Knowledge Graphs in Multimodal Urban Spaces. SIGMUS uses Large Language Models (LLMs) to produce the necessary world knowledge for identifying relationships between incidents occurring in urban spaces and data from different modalities, allowing us to organize evidence and observations relevant to an incident without relying and human-encoded rules for relating multimodal sensory data with incidents. This organized knowledge is represented as a knowledge graph, organizing incidents, observations, and much more. We find that our system is able to produce reasonable connections between 5 different data sources (new article text, CCTV images, air quality, weather, and traffic measurements) and relevant incidents occurring at the same time and location.

cs.AI↗

arXiv2025

Benchmarking Spatiotemporal Reasoning in LLMs and Reasoning Models: Capabilities and Challenges

Spatiotemporal reasoning plays a key role in Cyber-Physical Systems (CPS). Despite advances in Large Language Models (LLMs) and Large Reasoning Models (LRMs), their capacity to reason about complex spatiotemporal signals remains underexplored. This paper proposes a hierarchical SpatioTemporal reAsoning benchmaRK, STARK, to systematically evaluate LLMs across three levels of reasoning complexity: state estimation (e.g., predicting field variables, localizing and tracking events in space and time), spatiotemporal reasoning over states (e.g., inferring spatial-temporal relationships), and world-knowledge-aware reasoning that integrates contextual and domain knowledge (e.g., intent prediction, landmark-aware navigation). We curate 26 distinct spatiotemporal tasks with diverse sensor modalities, comprising 14,552 challenges where models answer directly or by Python Code Interpreter. Evaluating 3 LRMs and 8 LLMs, we find LLMs achieve limited success in tasks requiring geometric reasoning (e.g., multilateration or triangulation), particularly as complexity increases. Surprisingly, LRMs show robust performance across tasks with various levels of difficulty, often competing or surpassing traditional first-principle-based methods. Our results show that in reasoning tasks requiring world knowledge, the performance gap between LLMs and LRMs narrows, with some LLMs even surpassing LRMs. However, the LRM o3 model continues to achieve leading performance across all evaluated tasks, a result attributed primarily to the larger size of the reasoning models. STARK motivates future innovations in model architectures and reasoning paradigms for intelligent CPS by providing a structured framework to identify limitations in the spatiotemporal reasoning of LLMs and LRMs.

cs.AI↗

arXiv2025

The Untapped Potential of Smart Charging: How EV Owners Can Save Money and Reduce Emissions Without Behavioral Change

The transportation sector is the single largest contributor to US emissions and the second largest globally. Electric vehicles (EVs) are expected to represent half of global car sales by 2035, emerging as a pivotal solution to reduce emissions and enhance grid flexibility. The electrification of buildings, manufacturing, and transportation is expected to grow electricity demand substantially over the next decade. Without effectively managed EV charging, EVs could strain energy grid infrastructure and increase electricity costs. Drawing on de-identified 2023 EV telematics data from Rivian Automotive, this study found that 72% of home charging commenced after the customer plugged in their vehicle regardless of utility time of use (TOU) tariffs or managed charging programs. In fewer than 26% of charging sessions in the sample, EV owners actively scheduled charging times to align or participate in utility tariffs or programs. With a majority of drivers concurrently plugged in during optimal charging periods yet not actively charging, the study identified an opportunity to reduce individual EV owner costs and carbon emissions through smarter charging habits without significant behavioral modifications or sacrifice in user preferences. By optimizing home charging schedules within existing plug-in and plug-out windows, the study suggests that EV owners can save an average of $140 annually and reduce the associated carbon emissions of charging their EV by as much as 28%.

eess.SY↗

arXiv2024

A Recipe for CAC: Mosaic-based Generalized Loss for Improved Class-Agnostic Counting

Class agnostic counting (CAC) is a vision task that can be used to count the total occurrence number of any given reference objects in the query image. The task is usually formulated as a density map estimation problem through similarity computation among a few image samples of the reference object and the query image. In this paper, we point out a severe issue of the existing CAC framework: Given a multi-class setting, models don't consider reference images and instead blindly match all dominant objects in the query image. Moreover, the current evaluation metrics and dataset cannot be used to faithfully assess the model's generalization performance and robustness. To this end, we discover that the combination of mosaic augmentation with generalized loss is essential for addressing the aforementioned issue of CAC models to count objects of majority (i.e. dominant objects) regardless of the references. Furthermore, we introduce a new evaluation protocol and metrics for resolving the problem behind the existing CAC evaluation scheme and better benchmarking CAC models in a more fair manner. Besides, extensive evaluation results demonstrate that our proposed recipe can consistently improve the performance of different CAC models. The code is available at https://github.com/littlepenguin89106/MGCAC.

cs.CV↗

arXiv2024

Understanding factors behind IoT privacy -- A user's perspective on RF sensors

While IoT sensors in physical spaces have provided utility and comfort in our lives, their instrumentation in private and personal spaces has led to growing concerns regarding privacy. The existing notion behind IoT privacy is that the sensors whose data can easily be understood and interpreted by humans (such as cameras) are more privacy-invasive than sensors that are not human-understandable, such as RF (radio-frequency) sensors. However, given recent advancements in machine learning, we can not only make sensitive inferences on RF data but also translate between modalities. Thus, the existing notions of privacy for IoT sensors need to be revisited. In this paper, our goal is to understand what factors affect the privacy notions of a non-expert user (someone who is not well-versed in privacy concepts). To this regard, we conduct an online study of 162 participants from the USA to find out what factors affect the privacy perception of a user regarding an RF-based device or a sensor. Our findings show that a user's perception of privacy not only depends upon the data collected by the sensor but also on the inferences that can be made on that data, familiarity with the device and its form factor as well as the control a user has over the device design and its data policies. When the data collected by the sensor is not human-interpretable, it is the inferences that can be made on the data and not the data itself that users care about when making informed decisions regarding device privacy.

cs.HC↗

arXiv2023

Zero-shot Building Attribute Extraction from Large-Scale Vision and Language Models

Existing building recognition methods, exemplified by BRAILS, utilize supervised learning to extract information from satellite and street-view images for classification and segmentation. However, each task module requires human-annotated data, hindering the scalability and robustness to regional variations and annotation imbalances. In response, we propose a new zero-shot workflow for building attribute extraction that utilizes large-scale vision and language models to mitigate reliance on external annotations. The proposed workflow contains two key components: image-level captioning and segment-level captioning for the building images based on the vocabularies pertinent to structural and civil engineering. These two components generate descriptive captions by computing feature representations of the image and the vocabularies, and facilitating a semantic match between the visual and textual representations. Consequently, our framework offers a promising avenue to enhance AI-driven captioning for building attribute extraction in the structural and civil engineering domains, ultimately reducing reliance on human annotations while bolstering performance and adaptability.

cs.CV↗

arXiv2019

Towards Real-Time Respiratory Motion Prediction based on Long Short-Term Memory Neural Networks

Radiation therapy of thoracic and abdominal tumors requires incorporating the respiratory motion into treatments. To precisely account for the patient respiratory motions and predict the respiratory signals, a generalized model for predictions of different types of respiratory motions is desired. The aim of this study is to explore the feasibility of developing a Long Short-Term Memory (LSTM)-based model for the respiratory signal prediction. To achieve that, 1703 sets of Real-Time Position Management data were collected from retrospective studies across three clinical institutions. These datasets were separated as the training, internal validity and external validity groups. 1187 datasets were used for model development and the remaining 516 datasets were used to test the generality power of the model. Furthermore, an exhaustive grid search was implemented to find the optimal hyper-parameters of the LSTM model. The hyper-parameters are the number of LSTM layers, the number of hidden units, the optimizer, the learning rate, the number of epochs, and the length of time lags. Our model achieved superior accuracy over conventional artificial neural network models: with a prediction window of 500ms, the LSTM model achieved an average relative Mean Absolute Error (MAE) of 0.037, an average Root Mean Square Error (RMSE) of 0.048, and a Maximum Error (ME) of 1.687 in the internal validity data, and an average relative MAE of 0.112, an average RMSE of 0.139 and an ME of 1.811 in the external validity data. Compared to the LSTM model trained with default hyper-parameters, the MAE of the optimized model results decreased by 20%, indicating the importance of tuning the hyper-parameters of LSTM models to obtain superior accuracy. This study demonstrates the potential of deep LSTM models for the respiratory signal prediction and illustrates the impacts of major hyper-parameters in LSTM models.

physics.med-ph↗

arXiv2014

Negative Capacitance in a Ferroelectric Capacitor

The Boltzmann distribution of electrons poses a fundamental barrier to lowering energy dissipation in conventional electronics, often termed as Boltzmann Tyranny. Negative capacitance in ferroelectric materials, which stems from the stored energy of phase transition, could provide a solution, but a direct measurement of negative capacitance has so far been elusive. Here we report the observation of negative capacitance in a thin, epitaxial ferroelectric film. When a voltage pulse is applied, the voltage across the ferroelectric capacitor is found to be decreasing with time-in exactly the opposite direction to which voltage for a regular capacitor should change. Analysis of this inductance-like behavior from a capacitor presents an unprecedented insight into the intrinsic energy profile of the ferroelectric material and could pave the way for completely new applications.

cond-mat.mes-hall↗