arXiv subjects

Ryan Li

Publications and source records attributed to Ryan Li.

17 recordsLinked to original sources

arXiv2026

A Theory-Guided LLM Pedagogical Agent for STEM+C Scaffolding Without Over-Reliance

LLM pedagogical agents are proliferating, yet recent findings have raised questions about their adherence to established theories of learning and, by extension, their educational value. Concerns regarding cognitive offloading, over-reliance, and "gaming" behaviors persist and remain largely unaddressed. In response, we developed Copa, an agentic, multi-agent, multimodal Collaborative Peer Agent for STEM+C learning. Copa is built on top of the Evidence-Decision-Feedback (EDF) framework, grounding its interactions in Social Cognitive Theory and Social Constructivism and promoting sense-making through adaptive, dialogic support rather than answer-seeking. In an authentic high school computational-modeling study (n=33 dyads), we demonstrate that Copa (1) supports students' confidence building and ability to verbalize conceptual understanding without causing dependence; and (2) provides adaptive feedback personalized to learners that is interpretable with respect to students' multimodal input data. These findings position theory-guided, multimodal LLM agents as a promising path toward classroom AI integration that amplifies students' reasoning rather than replacing it.

cs.MA↗

arXiv2026

A new mixture model for spatiotemporal exceedances with flexible tail dependence

We propose a new model and estimation framework for spatiotemporal streamflow exceedances above a threshold that flexibly captures asymptotic dependence and independence in the tail of the distribution. We model streamflow using a mixture of processes with spatial, temporal and spatiotemporal asymptotic dependence regimes. A censoring mechanism allows us to use only observations above a threshold to estimate marginal and joint probabilities of extreme events. As the likelihood is intractable, we use simulation-based inference powered by random forests to estimate model parameters from summary statistics of the data. Simulations and modeling of streamflow data from the U.S. Geological Survey illustrate the feasibility and practicality of our approach.

stat.ME↗

arXiv2025

Card Dealing Math

Various card tricks involve under-down dealing, where alternatively one card is placed under the deck and the next card is dealt. We study how the cards need to be prepared in the deck to be dealt in order. The order in which the $N$ cards are prepared defines a permutation. In this work, we analyze general dealing patterns, considering properties of the resulting permutations. We give recursive formulas for these permutations, their inverses, the final dealt card, and the dealing order of the first card. We discuss some particular examples of dealing patterns and conclude with an analysis of several existing and novel magic card tricks making use of dealing patterns. Our discussions involve 30 existing sequences in the OEIS, and we introduce 44 new sequences to that database.

math.NT↗

arXiv2025

OpenCUA: Open Foundations for Computer-Use Agents

Vision-language models have demonstrated impressive capabilities as computer-use agents (CUAs) capable of automating diverse computer tasks. As their commercial potential grows, critical details of the most capable CUA systems remain closed. As these agents will increasingly mediate digital interactions and execute consequential decisions on our behalf, the research community needs access to open CUA frameworks to study their capabilities, limitations, and risks. To bridge this gap, we propose OpenCUA, a comprehensive open-source framework for scaling CUA data and foundation models. Our framework consists of: (1) an annotation infrastructure that seamlessly captures human computer-use demonstrations; (2) AgentNet, the first large-scale computer-use task dataset spanning 3 operating systems and 200+ applications and websites; (3) a scalable pipeline that transforms demonstrations into state-action pairs with reflective long Chain-of-Thought reasoning that sustain robust performance gains as data scales. Our end-to-end agent models demonstrate strong performance across CUA benchmarks. In particular, OpenCUA-72B achieves an average success rate of 45.0% on OSWorld-Verified, establishing a new state-of-the-art (SOTA) among open-source models. Further analysis confirms that our approach generalizes well across domains and benefits significantly from increased test-time computation. We release our annotation tool, datasets, code, and models to build open foundations for further CUA research.

cs.AI↗

arXiv2025

BountyBench: Dollar Impact of AI Agent Attackers and Defenders on Real-World Cybersecurity Systems

AI agents have the potential to significantly alter the cybersecurity landscape. Here, we introduce the first framework to capture offensive and defensive cyber-capabilities in evolving real-world systems. Instantiating this framework with BountyBench, we set up 25 systems with complex, real-world codebases. To capture the vulnerability lifecycle, we define three task types: Detect (detecting a new vulnerability), Exploit (exploiting a given vulnerability), and Patch (patching a given vulnerability). For Detect, we construct a new success indicator, which is general across vulnerability types and provides localized evaluation. We manually set up the environment for each system, including installing packages, setting up server(s), and hydrating database(s). We add 40 bug bounties, which are vulnerabilities with monetary awards from \$10 to \$30,485, covering 9 of the OWASP Top 10 Risks. To modulate task difficulty, we devise a new strategy based on information to guide detection, interpolating from identifying a zero day to exploiting a given vulnerability. We evaluate 10 agents: Claude Code, OpenAI Codex CLI with o3-high and o4-mini, and custom agents with o3-high, GPT-4.1, Gemini 2.5 Pro Preview, Claude 3.7 Sonnet Thinking, Qwen3 235B A22B, Llama 4 Maverick, and DeepSeek-R1. Given up to three attempts, the top-performing agents are Codex CLI: o3-high (12.5% on Detect, mapping to \$3,720; 90% on Patch, mapping to \$14,152), Custom Agent: Claude 3.7 Sonnet Thinking (67.5% on Exploit), and Codex CLI: o4-mini (90% on Patch, mapping to \$14,422). Codex CLI: o3-high, Codex CLI: o4-mini, and Claude Code are more capable at defense, achieving higher Patch scores of 90%, 90%, and 87.5%, compared to Exploit scores of 47.5%, 32.5%, and 57.5% respectively; while the custom agents are relatively balanced between offense and defense, achieving Exploit scores of 17.5-67.5% and Patch scores of 25-60%.

cs.CR↗

arXiv2025

Automated Hypothesis Validation with Agentic Sequential Falsifications

Hypotheses are central to information acquisition, decision-making, and discovery. However, many real-world hypotheses are abstract, high-level statements that are difficult to validate directly. This challenge is further intensified by the rise of hypothesis generation from Large Language Models (LLMs), which are prone to hallucination and produce hypotheses in volumes that make manual validation impractical. Here we propose Popper, an agentic framework for rigorous automated validation of free-form hypotheses. Guided by Karl Popper's principle of falsification, Popper validates a hypothesis using LLM agents that design and execute falsification experiments targeting its measurable implications. A novel sequential testing framework ensures strict Type-I error control while actively gathering evidence from diverse observations, whether drawn from existing data or newly conducted procedures. We demonstrate Popper on six domains including biology, economics, and sociology. Popper delivers robust error control, high power, and scalability. Furthermore, compared to human scientists, Popper achieved comparable performance in validating complex biological hypotheses while reducing time by 10 folds, providing a scalable, rigorous solution for hypothesis validation.

cs.LG↗

arXiv2024

FedGAT: A Privacy-Preserving Federated Approximation Algorithm for Graph Attention Networks

Federated training methods have gained popularity for graph learning with applications including friendship graphs of social media sites and customer-merchant interaction graphs of huge online marketplaces. However, privacy regulations often require locally generated data to be stored on local clients. The graph is then naturally partitioned across clients, with no client permitted access to information stored on another. Cross-client edges arise naturally in such cases and present an interesting challenge to federated training methods, as training a graph model at one client requires feature information of nodes on the other end of cross-client edges. Attempting to retain such edges often incurs significant communication overhead, and dropping them altogether reduces model performance. In simpler models such as Graph Convolutional Networks, this can be fixed by communicating a limited amount of feature information across clients before training, but GATs (Graph Attention Networks) require additional information that cannot be pre-communicated, as it changes from training round to round. We introduce the Federated Graph Attention Network (FedGAT) algorithm for semi-supervised node classification, which approximates the behavior of GATs with provable bounds on the approximation error. FedGAT requires only one pre-training communication round, significantly reducing the communication overhead for federated GAT training. We then analyze the error in the approximation and examine the communication overhead and computational complexity of the algorithm. Experiments show that FedGAT achieves nearly the same accuracy as a GAT model in a centralised setting, and its performance is robust to the number of clients as well as data distribution.

cs.LG↗

arXiv2024

Sketch2Code: Evaluating Vision-Language Models for Interactive Web Design Prototyping

Sketches are a natural and accessible medium for UI designers to conceptualize early-stage ideas. However, existing research on UI/UX automation often requires high-fidelity inputs like Figma designs or detailed screenshots, limiting accessibility and impeding efficient design iteration. To bridge this gap, we introduce Sketch2Code, a benchmark that evaluates state-of-the-art Vision Language Models (VLMs) on automating the conversion of rudimentary sketches into webpage prototypes. Beyond end-to-end benchmarking, Sketch2Code supports interactive agent evaluation that mimics real-world design workflows, where a VLM-based agent iteratively refines its generations by communicating with a simulated user, either passively receiving feedback instructions or proactively asking clarification questions. We comprehensively analyze ten commercial and open-source models, showing that Sketch2Code is challenging for existing VLMs; even the most capable models struggle to accurately interpret sketches and formulate effective questions that lead to steady improvement. Nevertheless, a user study with UI/UX experts reveals a significant preference for proactive question-asking over passive feedback reception, highlighting the need to develop more effective paradigms for multi-turn conversational agents.

cs.CL↗

arXiv2024

Semantic Self-Consistency: Enhancing Language Model Reasoning via Semantic Weighting

While large language models (LLMs) have rapidly improved their performance on a broad number of tasks, they still often fall short on reasoning tasks. As LLMs become more integrated in diverse real-world tasks, advancing their reasoning capabilities is crucial to their effectiveness in nuanced, complex problems. Wang et al.'s self-consistency framework reveals that sampling multiple rationales before taking a majority vote reliably improves model performance across various closed-answer reasoning tasks. Standard methods based on this framework aggregate the final decisions of these rationales but fail to utilize the semantic information detailed in the step-by-step reasoning paths. Our work introduces semantic self-consistency, enhancing this approach by incorporating and analyzing both the reasoning paths of these rationales in addition to their final decisions before taking a majority vote. These methods not only improve the reliability of reasoning paths but also cause more robust performance on complex reasoning tasks.

cs.CL↗

arXiv2024

CultureBank: An Online Community-Driven Knowledge Base Towards Culturally Aware Language Technologies

To enhance language models' cultural awareness, we design a generalizable pipeline to construct cultural knowledge bases from different online communities on a massive scale. With the pipeline, we construct CultureBank, a knowledge base built upon users' self-narratives with 12K cultural descriptors sourced from TikTok and 11K from Reddit. Unlike previous cultural knowledge resources, CultureBank contains diverse views on cultural descriptors to allow flexible interpretation of cultural knowledge, and contextualized cultural scenarios to help grounded evaluation. With CultureBank, we evaluate different LLMs' cultural awareness, and identify areas for improvement. We also fine-tune a language model on CultureBank: experiments show that it achieves better performances on two downstream cultural tasks in a zero-shot setting. Finally, we offer recommendations based on our findings for future culturally aware language technologies. The project page is https://culturebank.github.io . The code and model is at https://github.com/SALT-NLP/CultureBank . The released CultureBank dataset is at https://huggingface.co/datasets/SALT-NLP/CultureBank .

cs.CL↗

arXiv2024

Design2Code: Benchmarking Multimodal Code Generation for Automated Front-End Engineering

Generative AI has made rapid advancements in recent years, achieving unprecedented capabilities in multimodal understanding and code generation. This can enable a new paradigm of front-end development in which multimodal large language models (MLLMs) directly convert visual designs into code implementations. In this work, we construct Design2Code - the first real-world benchmark for this task. Specifically, we manually curate 484 diverse real-world webpages as test cases and develop a set of automatic evaluation metrics to assess how well current multimodal LLMs can generate the code implementations that directly render into the given reference webpages, given the screenshots as input. We also complement automatic metrics with comprehensive human evaluations to validate the performance ranking. To rigorously benchmark MLLMs, we test various multimodal prompting methods on frontier models such as GPT-4o, GPT-4V, Gemini, and Claude. Our fine-grained break-down metrics indicate that models mostly lag in recalling visual elements from the input webpages and generating correct layout designs.

cs.CL↗

arXiv2024

Beyond-classical computation in quantum simulation

Quantum computers hold the promise of solving certain problems that lie beyond the reach of conventional computers. However, establishing this capability, especially for impactful and meaningful problems, remains a central challenge. Here, we show that superconducting quantum annealing processors can rapidly generate samples in close agreement with solutions of the Schr\"odinger equation. We demonstrate area-law scaling of entanglement in the model quench dynamics of two-, three-, and infinite-dimensional spin glasses, supporting the observed stretched-exponential scaling of effort for matrix-product-state approaches. We show that several leading approximate methods based on tensor networks and neural networks cannot achieve the same accuracy as the quantum annealer within a reasonable time frame. Thus, quantum annealers can answer questions of practical importance that may remain out of reach for classical computation.

quant-ph↗

arXiv2023

Misinformation as a harm: structured approaches for fact-checking prioritization

In this work, we examine how fact-checkers prioritize which claims to fact-check and what tools may assist them in their efforts. Through a series of interviews with 23 professional fact-checkers from around the world, we validate that harm assessment is a central component of how fact-checkers triage their work. We also clarify the processes behind fact-checking prioritization, finding that they are typically ad hoc, and gather suggestions for tools that could help with these processes. To address the needs articulated by fact-checkers, we present a structured framework of questions to help fact-checkers negotiate the priority of claims through assessing potential harms. Our FABLE Framework of Misinformation Harms incorporates five dimensions of magnitude -- (social) Fragmentation, Actionability, Believability, Likelihood of spread, and Exploitativeness -- that can help determine the potential urgency of a specific message or claim when considering misinformation as harm. The result is a practical and conceptual tool to support fact-checkers and others as they make strategic decisions to prioritize their efforts. We conclude with a discussion of computational approaches to support structured prioritization, as well as applications beyond fact-checking to content moderation and curation.

cs.HC↗

arXiv2022

NL2INTERFACE: Interactive Visualization Interface Generation from Natural Language Queries

We develop NL2INTERFACE to explore the potential of generating usable interactive multi-visualization interfaces from natural language queries. With NL2INTERFACE, users can directly write natural language queries to automatically generate a fully interactive multi-visualization interface without any extra effort of learning a tool or programming language. Further, users can interact with the interfaces to easily transform the data and quickly see the results in the visualizations.

cs.HC↗

arXiv2022

Precipitation efficiency amplifies climate sensitivity by enhancing tropical circulation slowdown and eastern pacific warming

Cloud processes are the largest source of uncertainty in quantifying the global temperature response to carbon dioxide rise. Still, the role of precipitation efficiency (PE) -- surface rain per unit column -- integrated condensation -- is yet to be quantified. Here we use 36 limited-domain cloud resolving simulations from the Radiative-Convective Equilibrium Model Intercomparison Project to show that they strongly imply climate warming will result in increases to net precipitation efficiency. We then analyze 35 General Circulation Models (GCMs) from the Coupled Model Intercomparison Project Phase 6 and find that increasing PE enhances tropical circulation slowdown and strengthens eastern equatorial Pacific warming. These changes trigger pan-tropical positive cloud feedback by causing stratiform anvil cloud reduction and stratocumulus suppression, and thereby amplify overall climate sensitivity. Quantitatively, we find that in the 24 of 35 GCMs which match the cloud-resolving models in simulating increasing PE with greenhouse warming, mean Effective Climate Sensitivity is 1 K higher than in GCMs in which PE decreases. The models simulating increasing PE also comprise all estimates of effective climate sensitivity over 4 K. Taken together, these results show that further constraining PE sensitivity to warming will reduce uncertainty over future climate change.

physics.ao-ph↗

arXiv2021

Architectural considerations in the design of a third-generation superconducting quantum annealing processor

Early generations of superconducting quantum annealing processors have provided a valuable platform for studying the performance of a scalable quantum computing technology. These studies have directly informed our approach to the design of the next-generation processor. Our design priorities for this generation include an increase in per-qubit connectivity, a problem Hamiltonian energy scale similar to previous generations, reduced Hamiltonian specification errors, and an increase in the processor scale that also leaves programming and readout times fixed or reduced. Here we discuss the specific innovations that resulted in a processor architecture that satisfies these design priorities.

quant-ph↗

arXiv2019

Scaling advantage in quantum simulation of geometrically frustrated magnets

The promise of quantum computing lies in harnessing programmable quantum devices for practical applications such as efficient simulation of quantum materials and condensed matter systems. One important task is the simulation of geometrically frustrated magnets in which topological phenomena can emerge from competition between quantum and thermal fluctuations. Here we report on experimental observations of relaxation in such simulations, measured on up to 1440 qubits with microsecond resolution. By initializing the system in a state with topological obstruction, we observe quantum annealing (QA) relaxation timescales in excess of one microsecond. Measurements indicate a dynamical advantage in the quantum simulation over the classical approach of path-integral Monte Carlo (PIMC) fixed-Hamiltonian relaxation with multiqubit cluster updates. The advantage increases with both system size and inverse temperature, exceeding a million-fold speedup over a CPU. This is an important piece of experimental evidence that in general, PIMC does not mimic QA dynamics for stoquastic Hamiltonians. The observed scaling advantage, for simulation of frustrated magnetism in quantum condensed matter, demonstrates that near-term quantum devices can be used to accelerate computational tasks of practical relevance.

quant-ph↗