Searcharxiv⌕ Search

arXiv subjects

Xinyi Chen

Publications and source records attributed to Xinyi Chen.

At least 19 recordsLinked to original sources
arXiv2026

Evo-DexBench: Benchmarking Generalization across Related Tasks for Dexterous Manipulation

General-purpose dexterous manipulation requires policies to transfer knowledge from related tasks to new tasks with only a few target demonstrations. Existing benchmarks often provide too many target demonstrations and leave the differences between source and target tasks uncontrolled. As a result, they cannot clearly distinguish knowledge transfer from task memorization. Our key insight is the Core-Gen principle. Core tasks are related tasks that share a functional interaction and provide source knowledge, while a Gen task is a novel target within the same functional family that has only a few demonstrations. This setup tests whether a policy can reuse functional interaction knowledge. Based on this principle, we propose Evo-DexBench, which controls both source-task relatedness and target-task exposure. It comprises 30 tasks in six functional families, three everyday scenes per task, and 1,248 multimodal demonstrations collected via VR teleoperation with simulated Shadow Hands equipped with 123 taxels. Experiments show that, among six representative policies, even the best-performing method achieves an average success rate of only 24.4%, highlighting the challenges existing methods face in transferring functional interaction knowledge. Further analysis shows that, across the two tasks, joint Core + Gen training improves performance by 12% and 38%, respectively, over training on Core followed by fine-tuning on Gen, indicating that joint training makes better use of Core knowledge. With matched Core data, 8 Gen demonstrations achieve performance comparable to training on 16--32 Gen demonstrations alone. On all three tested Gen targets, functionally matched Core data yields greater gains in success rate than mismatched data. We release the Evo-DexBench dataset and benchmark to support the evaluation of dexterous manipulation policies' generalization across related tasks.

cs.RO↗
arXiv2026

Measurement of the galaxy-velocity power spectrum of DESI tracers with the kinematic Sunyaev-Zeldovich effect using DESI DR2 and ACT DR6

Joint analyses of high-resolution CMB temperature maps with galaxy surveys provide a unique way to reconstruct the radial velocity field of the underlying matter distribution via the kinematic Sunyaev-Zeldovich (kSZ) effect. Using data from the Atacama Cosmology Telescope (ACT) DR6 and the Dark Energy Spectroscopic Instrument (DESI) DR2, we present radial velocity reconstructions for luminous red galaxies (LRGs), emission-line galaxies (ELGs), and quasars (QSOs). Leveraging the spectroscopic data, we are able to reliably model the foreground contamination and report a negligible impact on our main observables. We detect the velocity-galaxy cross-correlation at $17.0σ$ for LRGs, and for the first time, at $8.3σ$ for ELGs and $6.8σ$ for QSOs. We further report the first detection of the velocity-velocity correlation using LRGs at $3.1σ$, as well as the highest cumulative detection of the kSZ effect to date at $20.8 σ$. Similarly to previous results, we find a lower amplitude of the kSZ signal compared to our fiducial halo model prediction and electron profile assuming a Battaglia profile. Combining these new observables, we obtain constraints on local-type primordial non-Gaussianity (PNG): $f_{\rm NL}^{\rm loc} = 15.9_{-34.4}^{+34.6}$ at 68\% confidence, which represents the tightest constraint to date derived from the velocity field. The measurements presented here already exhibit lower noise on a per-mode basis than the galaxy auto-correlation on the largest scales, $k<0.004~\rm{Mpc^{-1}}$, highlighting the key role these observables will play in the context of future CMB experiments such as the Simons Observatory.

astro-ph.CO↗
arXiv2026

RAVEN: A Unified Active SLAM System for Rapid Autonomous Visual Exploration with Aerial Robots under Localization Uncertainty

Active simultaneous localization and mapping (SLAM) addresses the problem of planning robot motion to build complete and accurate maps of unknown environments through efficient exploration while maintaining high localization accuracy. This raises a dilemma between exploring unknown spaces to expand coverage and exploiting visited regions to perform active loop closures. Despite numerous prior works on ground robots, aerial robots with high agility and constrained sensor suites face more significant localization drift and thereby demand dedicated active SLAM planning. Moreover, a unified strategy that adapts performance based on task-specific priorities remains an open question. In this article, we introduce RAVEN, a unified active SLAM framework for Rapid Autonomous Visual ExploratioN using quadrotors, which systematically accounts for localization uncertainty. The framework starts with identifying frontiers and active loop closure candidates, from which global tour candidates are generated. These candidates are evaluated using a utility vector that separately quantifies exploration efficiency and localization accuracy, and the final tour is selected according to a prioritization parameter ranging from $[0,1]$. Finally, a perception-aware navigation planner generates smooth execution trajectories that minimize visual state estimation error based on the online Fisher information field that provides visual localization information in real-time. Benchmark evaluations demonstrate that RAVEN outperforms state-of-the-art active SLAM methods, highlighting its advantages in producing a spectrum of performance across prioritizations. Fully onboard field experiments with a customized stereo-inertial quadrotor further validate its practical applicability in challenging environments. The source code will be released to benefit the community.

cs.RO↗
arXiv2026

Novel or Not? PatEx-Gym: Agentic Prior-Art Search for Novelty Assessment in Patent Examination

Patent examination is a challenging real-world task that requires agents to discover sparse prior art, iteratively refine queries, inspect technical evidence, and produce novelty assessments documented in Office Actions (OAs) across heterogeneous scientific and technical domains. However, existing benchmarks largely focus on static and limited candidate sets, leaving the full search--reasoning--assessment process underexplored. We present PatEx-Gym, the first framework for developing and evaluating agents for patent examination, which integrates three components. First, PatEx-Bench provides 1K realistic examination tasks, two large-scale retrieval corpora containing over 1M patents and 52K scientific papers, and a harness runtime supporting iterative search, document visits, reasoning, and OA generation. Second, we introduce PASS, an annotation-free Prior-Art Agentic Search Trajectory Synthesis framework. Specifically, it exploits document visits as implicit relevance signals and evidence-acquisition anchors to model intra-turn query behavior, inter-turn search progression, and trajectory-level evidence and outcome quality, yielding PatEx-SynData. Third, we train PatEx-Agent at the 4B scale via supervised fine-tuning on the curated trajectories, establishing a lightweight baseline. Extensive experiments show that final decision accuracy alone does not fully reflect examination quality, while prior-art discovery remains difficult even for strong frontier models. Our PatEx-Agent achieves competitive performance against substantially larger general-purpose agents. These results show that training on PatEx-SynData enables a compact 4B model to serve as a competitive baseline for prior-art search and patent examination. We view patent examination as a demanding testbed for agentic search and deep research agents.

cs.IR↗
arXiv2026

EvalCards: A Framework for Standardized Evaluation Reporting

Evaluation reports guide the use of general-purpose AI (GPAI) models, yet inconsistent reporting often leaves essential questions unanswered. Through a review of prior work, we identify three central issues in evaluation reporting practices today, reproducibility, accessibility, and accountability, and examine how existing reporting artifacts fall short in addressing them. We introduce the EvalCards framework, outlining its key dimensions and presenting a preliminary format for concise evaluation summaries. We conclude by discussing how the framework can guide the future development of evaluation reporting standards.

cs.CL↗
arXiv2026

Observability from measurable sets for strongly coupled parabolic systems

We characterize observability of strongly coupled parabolic systems from arbitrary space-time measurable sets of positive measure. For constant coercive diffusion matrices and arbitrary constant zeroth-order coupling and observation matrices, observability holds if and only if the generalized Kalman rank condition is satisfied at every Dirichlet eigenvalue. The diffusion matrix need not be diagonalizable, and the number of components is arbitrary. The observation is measured in $L^1$, yielding null controllability with bounded controls by duality. The proof combines a Remez-type inequality for exponential polynomials with a quantitative analysis of the mode-dependent Kalman condition. The resulting finite-dimensional observability estimate is uniform over the spatial modes up to a polynomial frequency loss. A spectral inequality and the Lebeau--Robbiano strategy then give observability of the full system.

math.OC↗
arXiv2026

PRISM: Prior Relational Information for Self-supervised Modeling to Enhance Solubility OOD Generalization

Solubility depends on intermolecular interactions between solutes and solvents, and is central to drug discovery, chemical and materials process engineering. Its prediction requires out-of-distribution (OOD) generalization to unseen solutes, solvents, and solvent mixtures, where supervised methods suffer from data scarcity, and existing self-supervised methods do not explicitly learn chemically meaningful relations. Learning chemically meaningful relational information without labels can alleviate data scarcity and improve OOD generalization, but remains challenging. We introduce Prior Relational Information for Self-supervised Modeling (PRISM) and apply it to solubility OOD generalization. PRISM is a label-free pretraining framework that converts expert-defined chemical priors into relational teachers. For solubility prediction, it trains four encoders to preserve the population-level relational information from priors capturing hydrophobic, electrostatic, polarizability, and substructural effects, using Kullback-Leibler (KL) divergence to match distance-based neighborhood distributions during pretraining. Across scaffold, solute, and solvent OOD settings, PRISM matches or surpasses supervised and self-supervised baselines. Under mixed-solvent OOD, freezing the four PRISM encoders and training a lightweight fusion-and-composition head outperforms strong baselines and substantially reduces run-to-run variation. Ablations and control studies confirm that gains stem from chemically meaningful relational structures baked into the encoders by PRISM. We expect this strategy of turning relational priors into self-supervision signals to extend beyond solubility prediction to domains where chemically or biologically meaningful relational priors are available.

cs.LG↗
arXiv2026

Achieving Linear Speedup and Near-Optimal Complexity for Decentralized Optimization over Row-stochastic Networks

A key challenge in decentralized optimization is determining the optimal convergence rate and designing algorithms to achieve it. While this problem has been extensively addressed for doubly-stochastic and column-stochastic mixing matrices, the row-stochastic scenario remains unexplored. This paper bridges this gap by introducing effective metrics to capture the influence of row-stochastic mixing matrices and establishing the first convergence lower bound for decentralized learning over row-stochastic networks. However, existing algorithms fail to attain this lower bound due to two key issues: deviation in the descent direction caused by the adapted gradient tracking (GT) and instability introduced by the Pull-Diag protocol. To address descent deviation, we propose a novel analysis framework demonstrating that Pull-Diag-GT achieves linear speedup, the first such result for row-stochastic decentralized optimization. Moreover, by incorporating a multi-step gossip (MG) protocol, we resolve the instability issue and attain the lower bound, achieving near-optimal complexity for decentralized optimization over row-stochastic networks.

math.OC↗
arXiv2026

SpikeAid: A Streaming Spiking Hearing Aid Framework for Joint Noise Reduction and Wide Dynamic Range Compression

Modern hearing aids combine noise reduction (NR) with wide dynamic range compression (WDRC) to suppress background interference while compensating for the reduced auditory dynamic range associated with hearing loss. Conventional systems typically process NR and WDRC as separate stages, although the output of NR directly affects the signal levels and envelopes that determine subsequent compression. Recent end-to-end neural approaches can jointly optimize these processes, but their dense computation remains challenging for always-on hearing aids. In this work, we propose SpikeAid, a causal end-to-end spiking neural network (SNN) framework for joint NR and WDRC. SpikeAid consists of an NR-Net that constructs complementary speech and noise representations and a WDRC-Net that estimates the corresponding branch gains, with both components jointly optimized according to the final hearing aid objective. To capture the multiscale temporal structures required by joint NR and WDRC in a computationally efficient manner, we further introduce the Delayed Resonate-and-fire Neuron (DRN), which incorporates bio-inspired heterogeneous dendritic delays and resonating dynamics. Experimental results on the hearing aid task show that SpikeAid improves key objective metrics over cascaded processing and outperforms existing spiking speech-processing baselines. Evaluation on the public Intel Neuromorphic Deep Noise Suppression (N-DNS) Challenge further demonstrates state-of-the-art neuromorphic speech-denoising performance with DRN. Finally, we designed an FPGA-based neuromorphic hardware system for efficient deployment of SpikeAid, where it achieves over two orders of magnitude higher energy efficiency than conventional CPU/GPU implementations, demonstrating its potential for energy-efficient hearing aid applications.

eess.AS↗
arXiv2026

MedMemoryBench: Benchmarking Agent Memory in Personalized Healthcare

The large-scale deployment of personalized healthcare agents demands memory mechanisms that are exceptionally precise, safe, and capable of long-term clinical tracking. However, existing benchmarks primarily focus on daily open-domain conversations, failing to capture the high-stakes complexity of real-world medical applications. Motivated by the stringent production requirements of an industry-leading health management agent serving tens of millions of active users, we introduce MedMemoryBench. We develop a human-agent collaborative pipeline to synthesize highly realistic, long-horizon medical trajectories based on clinically grounded, synthetic patient archetypes. This process yields a massive, expertly validated dataset comprising approximately 2,000 sessions and 16,000 interaction turns. Crucially, MedMemoryBench departs from traditional static evaluations by pioneering an "evaluate-while-constructing" streaming assessment protocol, which precisely mirrors dynamic memory accumulation in production environments. Furthermore, we formalize and systematically investigate the critical phenomenon of memory saturation, where sustained information influx actively degrades retrieval and reasoning robustness. Comprehensive benchmarking reveals severe bottlenecks in mainstream architectures, particularly concerning complex medical reasoning and noise resilience. By exposing these fundamental flaws, MedMemoryBench establishes a vital foundation for developing robust, production-ready medical agents.

cs.AI↗
arXiv2026

ComManip: Overfitting Manipulation Policies to Comfortable Regions

Training robot manipulation policies relies on costly robot demonstrations, making large-scale data collection impractical. Meanwhile, to improve policy generalization, existing approaches seek greater diversity in visual observations by varying object placements, viewpoints, and robot configurations during data collection. However, under a limited demonstration budget, this strategy forces the policy to model diverse visual observations, providing insufficient supervision to learn reliable observation-action correspondences under similar local conditions. Our study reveals that policies trained under this strategy achieve lower task success rates than those trained within a compact, visually and kinematically stable region. We refer to these stable regions as comfortable manipulation regions. To exploit this finding, we propose ComManip, a learning paradigm that specializes manipulation policies to comfortable manipulation regions. During inference, ComManip repositions the mobile base until the detected target center enters the familiar image-space range estimated from comfortable-region demonstrations. It then executes the same manipulation policy, enabling effective manipulation across diverse target locations. We conduct extensive experiments across multiple manipulation tasks, demonstration budgets, and policy families including ACT, $π_{0.5}$, RDT, OpenVLA-OFT, and SmolVLA. The results demonstrate that ComManip improves task success by roughly 20 percentage points or more across different policy architectures in large workspaces under limited demonstration budgets, suggesting that specializing manipulation policies to comfortable regions provides a more data-efficient learning paradigm for manipulation.

cs.RO↗
arXiv2026

Connecting the Dots: Prospects for Detecting Multi-Source Inflation through Scale-Dependent and Stochastic Bias of Little Red Dots

Recent work has proposed using the spatial clustering of Little Red Dots (LRDs) as a high-redshift cosmological probe. In this paper, we investigate the feasibility of obtaining a LRD sample with a future space experiment, and forecast LRD constraints on primordial local-type non-Gaussianity, $f^{\rm loc}_{\rm NL}$, and the collapsed-limit trispectrum amplitude, $τ^{\rm loc}_{\rm NL}$. LRDs are ideally suited to probe these inflationary parameters because they span a large cosmological volume, have high bias and are easily identifiable. We perform idealised joint $f_{\rm NL}^{\rm loc}$--$τ_{\rm NL}^{\rm loc}$ Fisher forecasts assuming the universality relation, Gaussian covariance, Poisson shot-noise and a normalised redshift dispersion, $σ(z) / \left( {1+z} \right)= 0.03$. We quantify the significance at which one would detect multi-source inflation under the Suyama--Yamaguchi relation, $τ^{\rm loc}_{\rm NL} > \left( 6/5 f_{\rm NL}\right)^2$, across the $f_{\rm NL}^{\rm loc}$--$τ_{\rm NL}^{\rm loc}$ plane. We find that a LRD analysis could robustly detect multi-source inflation at high-significance over a large region of parameter space, and would be more constraining than all other past or present experiments.

astro-ph.CO↗
arXiv2026

VideoTok4D: A 4D-Aware Video Tokenizer for Compact World Representation

Video tokenizers have emerged as a cornerstone of modern video modeling, underpinning progress in compression, reconstruction and generation by mapping high-dimensional visual signals into compact latent spaces. However, despite this progress, current tokenization paradigms largely remain within the 2D visual domain, treating videos as image sequences rather than observations of an underlying dynamic 3D world. Consequently, the learned tokens inherit this observation-centric bias, limiting their capacity to compactly represent real-world 4D scenes. To mitigate this issue, we propose VideoTok4D, a novel 4D-aware video tokenizer for compact world representation. Specifically, our approach comprises three key designs: 1) a spatiotemporal disentanglement strategy that factorizes videos into static and dynamic tokens for holistic world modeling; 2) a track-aware dynamic attention mechanism that aggregates trajectory-aligned cues to promote cross-view motion consistency; and 3) Co4DGen, a diffusion prior learned over the resulting VideoTok4D token space for efficient 4D scene generation. Extensive experiments have demonstrated that our proposed method achieves state-of-the-art performance while requiring up to 4 orders of magnitude less storage than dense 4D representations. Moreover, the compact token space substantially shortens diffusion sequences, enabling efficient generation.

cs.CV↗
arXiv2026

JudgmentLens: Human-AI Sensemaking of Complex Legal Judgments

Judicial judgments are increasingly available, yet dense language and distributed relationships among facts, evidence, reasoning, and rulings remain difficult for non-experts to interpret. Through a mixed-methods formative study with Chinese non-expert readers (survey N=34; interviews N=6), we identified structural, interpretive, verification, and action breakdowns. We developed JudgmentLens, an AI-augmented reading system combining persistent case representations, adaptive explanations, and traceable links from generated interpretations to judgment passages. In a counterbalanced within-subject evaluation (N=16), participants completed tasks faster with JudgmentLens than with conventional PDF reading and reported lower workload and greater self-reported decision understanding, while rubric-scored comprehension did not differ reliably. An exploratory PDF+DeepSeek probe suggested that conversational AI supported formulated questions while leaving question formulation, answer integration, and source checking largely to users. We contribute an empirical account of non-expert judgment sensemaking and design strategies for inspectable, source-grounded AI mediation.

cs.HC↗
arXiv2026

NetConfArena: An Executable Benchmark for LLM Agents in Closed-Loop Network Configuration

Large language model (LLM) agents are increasingly attractive for automating network configuration, yet their reliability and failure patterns are poorly understood. An essential prerequisite is to assess such agents in a realistic but risk-free environment. Existing benchmarks, however, fall short: they often treat configuration as static command generation or rely on overly simplified settings. Such evaluations understate the core challenges of network configuration, where correctness requires reasoning about protocol complexity and topology dependence. We present NetConfArena, an executable benchmark for evaluating LLM agents in closed-loop network configuration. NetConfArena places agents in emulated multi-device networks, provides a standardized and compact action interface for task execution, and evaluates the resulting network behavior with hidden task-specific executable test cases. The benchmark relies on an LLM-assisted, emulation-grounded pipeline, which converts human-oriented network materials into reusable parameterized task templates. We evaluate representative LLM agents on 480 task instances instantiated from 96 protocol-focused task templates, yielding 3840 execution trajectories, and show that failures are not limited to command errors. The failures also reveal gaps in task-specification adherence and robust planning and execution. These findings suggest two future directions: using validated trajectories as supervision signals to improve foundation models, and designing harness mechanisms that make agent execution more reliable and accountable.

cs.NI↗
arXiv2026

Fractal Remez inequality on the sphere and observability of the heat equation

This paper is concerned with Remez-type inequalities and their applications in observability inequality. Our aim is twofold. First, we establish the following fractal Remez's inequality on the unit sphere $\mathbb{S}^{n-1}$ \begin{align*} \sup_{\mathbb{S}^{n-1}} |p|\le C(M,N,n,δ)\sup_{M} |p|, \end{align*} where $M \subset \mathbb{S}^{n-1}$ ($n \ge 2$) is a fractal set of positive $(n-2+δ)$-Hausdorff content for arbitrary $δ\in (0,1)$, and $p$ is a spherical polynomial of degree at most $N\in \mathbb{Z}^+$. Second, building upon this fractal framework, we establish sharp observability inequalities for the heat equation on the sphere, again valid for all $δ\in (0, 1)$, which improve the result of Burq and Moyano [J. Eur. Math. Soc. (JEMS), 25 (4) (2023)] in the spherical setting. Furthermore, as an additional application, we prove a lower-dimensional observability inequality for the heat equation with super-quadratic potentials $V(x) = |x|^{2m}$ ($m \in \mathbb{Z}^+, m\ge 2$) on the whole space $\mathbb{R}^n$.

math.AP↗
arXiv2026

PMSN: A Parallel Multi-compartment Spiking Neuron for Multi-scale Temporal Processing

Spiking Neural Networks (SNNs) hold great potential to realize brain-inspired, energy-efficient computational systems. However, current SNNs still fall short in terms of multiscale temporal processing compared to their biological counterparts. This limitation has resulted in poor performance in many pattern recognition tasks with information that varies across different timescales. To address this issue, we put forward a novel spiking neuron model called Parallel Multi-compartment Spiking Neuron (PMSN). The PMSN emulates biological neurons by incorporating multiple interacting substructures and allows for flexible adjustment of the substructure counts to effectively represent temporal information across diverse timescales. Additionally, to address the computational burden associated with the increased complexity of the proposed model, we introduce two parallelization techniques that decouple the temporal dependencies of neuronal updates, enabling parallelized training across different time steps. Our experiments across a wide range of pattern recognition tasks demonstrate that PMSN outperforms state-of-the-art spiking neuron models in temporal processing capacity and training speed. Specifically, compared with the commonly used Leaky Integrate-and-Fire neuron, PMSN offers more than 10 times acceleration and a 30% accuracy improvement on Sequential CIFAR-10 dataset, while maintaining comparable computational cost. Our implementation on neuromorphic hardware further demonstrates the deployability of PMSN and highlights its favorable trade-off between effectiveness and efficiency. Therefore, the proposed PMSN presents a promising solution to harness the computational advantages of detailed biological neurons, enabling high-performance and efficient temporal processing on neuromorphic computing systems. Code is available at https://github.com/xychen-comp/PMSN.

cs.NE↗
arXiv2026

Visored: A Controlled-Natural-Language Prover for LLM-Generated Mathematics

We present a dependent-type-based prover designed around the way LLMs (and humans) tend to write mathematics, complementing existing systems such as Lean and Rocq. Its core design choices are a surface that imitates mathematical natural language and a rule-driven automation layer that closes the routine steps a textbook would omit, so that an accepted proof can be re-emitted as a checked Lean file. Early experiments suggest that, even without any prover-specific training data, LLMs can learn to use it effectively on the miniF2F benchmark. Lean output excerpts: https://github.com/xiyuzhai-husky-lang/visored/

cs.PL↗