arXiv subjects

Shuo Zhang

Publications and source records attributed to Shuo Zhang.

At least 19 recordsLinked to original sources

arXiv2026

Constant Approximation of Fréchet Distance in Near-Linear Time

We give the first constant-factor approximation for the Fréchet distance in near-linear time. Given two polygonal curves in $\mathbb{R}^d$ with $n$ and $m$ vertices, respectively, where $m\le n$, our deterministic algorithm runs in $O(n^{1+\varepsilon})$ time for every fixed $\varepsilon>0$ and $d$, with an approximation factor depending only on $\varepsilon$ and $d$. The same result holds for the discrete Fréchet distance. Our result extends the line of work initiated by Cheng, Huang, and Zhang (STOC 2025) from strongly subquadratic to near-linear time. Our algorithm recursively balances preprocessing and query time. An index for comparing subcurves of a single curve supports faster queries between different curves. We obtain this speedup by replacing one curve with surrogates assembled from indexed subcurves of the other and shared across starting points of matchings. Faster queries then reduce the cost of building the next index. Compact boundary reachability summaries allow the answers to be reused in later recursive calls. After a number of rounds depending only on $\varepsilon$, both preprocessing and query time are near-linear.

cs.CG↗

arXiv2026

NorthCUA Technical Report: Turbocharging Computer-Use Agents through Experience at Scale

Scaling computer-use agents from individual actions to complex software work requires experience that connects decisions to complete outcomes. We present NorthCUA, an integrated system for turbocharging computer-use agents through experience at scale. Its Experience Flywheel supplies diverse executable tasks to the NorthCUA Harness, which sustains long interactions, records decisions, and checks task outcomes. The Flywheel uses these records to curate verified trajectories. For unsolved tasks, recursive self-improvement evolves reusable skills and preserves verified milestones so later attempts build on earlier progress. Staged Learning combines this experience with grounding data from the Flywheel to develop GUI competence through mid-training and long-horizon supervised fine-tuning. The policy then learns from fresh attempts in the same Harness through joint desktop-web reinforcement learning, closing the loop between experience production and policy improvement. The resulting Nano, Mini, and Pro family achieves state-of-the-art results among open-source models across GUI grounding, desktop execution, and Web interaction under the reported evaluation settings. For more responsive Web workflows, NorthCUA-Ultrafast pairs direct action selection with open-ended generation on pre-RL Nano, offering a measured quality-latency tradeoff. Together, these advances establish NorthCUA as a high-performing model family and a reusable foundation for agents that carry complex software work from intent to completion.

cs.LG↗

arXiv2026

Quantum identity authentication protocol via perturbed probability amplitudes

Quantum identity authentication is an effective solution for establishing mutual trust between communicating parties in the post-quantum era. To address the risk of partial secret information leakage under special measurements, this paper designs a quantum identity authentication protocol based on perturbed probability amplitudes. The protocol first proposes a method of encoding a random sequence onto the probability amplitudes of quantum entangled states, forming perturbed probability amplitudes, thereby reducing the involvement of shared information between the sender and receiver. Furthermore, it presents a method for generating n-bit random authentication information based on the n-bit pre-shared information, enhancing the randomness of the authentication information relative to an attacker. The proposed protocol can resist impersonation attack, CNOT attack, and entanglement measurement attack, and ensures backward security. Under equivalent security conditions, the amount of shared information is 1/n that in Chen et al.'s scheme and half that in Dutta et al.'s scheme. And each bit of the secret message has no risk of leakage under special measurements. It can also be extended to a bi-directional quantum identity authentication protocol.

quant-ph↗

arXiv2026

HAWC, Fermi-LAT and Chandra Detections of High Energy Particles in the S106 Star Forming Region

Local star-forming regions have recently been the subject of scrutiny as cosmic-ray accelerators and sources of ultra-high-energy gamma-ray emissions. The Cygnus-X region is one of the richest star forming regions in the galaxy and over the last few decades has been extensively studied. Here we report the detection of a star-forming region located in the southern part of Cygnus-X, Sharpless 106 (S106), using the HAWC Observatory. The S106 region is spatially located near the cataloged gamma-ray source 1LHAASO J2027+3657. In this work, we perform a multi-wavelength analysis of S106, by combining HAWC data with re-analysis of Fermi-LAT and archival Chandra data. We construct a fully leptonic model for the multi-wavelength emission from S106 from a single population of high-energy leptons, with the gamma-rays greater than 1 TeV being produced by inverse Compton scattering of the locally-produced IR photon field from S106 itself. Our models show that hadronic mechanisms are not always required to explain the observed flux from gamma-ray star-forming regions and that understanding the specific details of local photon fields is crucial to modeling the multi-wavelength emission from star-forming regions.

astro-ph.HE↗

arXiv2026

Where Prefill Expert Reduction Hurts MoE Models: A Few Prompt Tokens Carry Most of the Short-Answer F1 Loss

Running prompt tokens through fewer routed experts during prefill makes mixture-of-experts (MoE) serving cheaper, but what it damages has not been measured. We locate the damage with token-, layer- and attention-path interventions in four MoE families on short-answer question answering. At half the experts the F1 loss is mostly answer form: the model still finds the answer but no longer answers briefly. The loss is made at a few tokens: the two chat-template tokens before the answer in Qwen3-30B-A3B, and the first 8-32 tokens, where the instruction sits, in DeepSeek-V2-Lite. Keeping them at full experts removes most of the F1 loss and about half and a fifth of the increase in answer NLL. These tokens are not the familiar suspects: reduction leaves the first-token sink within 4%, and restoring super experts or massive activations recovers at most a quarter of the loss. On real prompts the localisation becomes a placement rule: with about 4% of tokens at full experts, putting them on the task interface keeps F1 within one point of the unmodified model, while the same budget scattered in the documents is about 3 points worse in both prespecified primary tests of this comparison. Time to first token falls about 14% on 11K-token prompts.

cs.CL↗

arXiv2026

Pricing Information in Search with Nonobligatory Inspection

As AI agents reshape Web search and advertising, platforms can monetize information that improves decisions. We study information pricing for sequential search among independent alternatives with nonobligatory inspection: the buyer may select any alternative without buying its report. We show that a common fee maximizes static revenue for identical priors, resolving a uniform-pricing conjecture. Our results also establish that exact optimization is NP-hard for heterogeneous binary priors. Despite this hardness, our polynomial-time static pricing algorithm achieves a $3$-approximation to the information surplus. More generally, a common static fee approximates optimal adaptive revenue within a logarithmic factor in the number of alternatives. This guarantee is asymptotically tight.

cs.GT↗

arXiv2026

Piecewise $P_2$ finite element space on rectangular meshes in arbitrary dimensions for the biharmonic equation

This paper develops a piecewise $P_2$ finite element space on rectangular meshes in arbitrary dimensions for the biharmonic equation. The discrete space imposes continuity of vertex sums on $(d-2)$-faces and of averaged normal derivatives across interior $(d-1)$-faces. A local algebraic characterization of $P_2$ is first established through the unisolvence of an enriched polynomial space and the resulting compatibility conditions. For each interior element, a $3^d$-element patch is then introduced, and the associated homogeneous patch space is shown to be one-dimensional. The corresponding patch functions, extended by zero, form a basis of the global homogeneous discrete space. A local quadratic interpolation operator is further constructed in arbitrary dimensions. Its approximation properties, together with a consistency estimate derived from the weak continuity conditions, lead to the optimal broken $H^2$ error estimate for the biharmonic problem.

math.NA↗

arXiv2026

Energy Discontinuity of the Lyapunov Exponent for Quasiperiodic Schrödinger Operators with Gevrey Potentials

We construct quasiperiodic Schrödinger operators with Gevrey potentials for which the Lyapunov exponent is discontinuous in the spectral energy. More precisely, let $α\in\R\setminus\mathbb Q$ satisfy the Diophantine condition $\|kα\|_{\T}\geqγ|k|^{-τ}$ for all $k\in\Z\setminus\{0\}$, with $γ>0$ and $τ\geq1$. For every Gevrey index $\mathfrak s>1+τ$ and every sufficiently large coupling $λ$, we construct a real-valued potential $v_λ\in G^{\mathfrak s}(\T,\R)$ and a sequence of energies $E_j\downarrow0$ such that \[ L_{v_λ}(0)\geq(1-\varepsilon_λ)\logλ, \qquad \sup_{j\geq1} L_{v_λ}(E_j)\leq(1-c_*)\logλ, \] where $L_{v_λ}(E)$ denotes the Lyapunov exponent of the Schrödinger operator with potential $λv_λ$, $c_*>0$ is an absolute constant, and $\varepsilon_λ\to0$ as $λ\to\infty$. For bounded-type frequencies ($τ=1$), the regularity range $\mathfrak s>2$ is sharp up to the endpoint $\mathfrak s=2$, since the Lyapunov exponent is known to be continuous in energy for $1<\mathfrak s<2$.

math.DS↗

arXiv2026

Dual-Channel Contrastive Encoding of Sperm Whale Coda

Sperm whales (Physeter macrocephalus) communicate through short click sequences called codas, whose structure includes the rhythmic pattern of inter-click intervals (ICI) and the spectral composition of the clicks. In this paper, we propose the Dual-Channel Contrastive Encoder (DCCE), which explicitly encodes rhythm and spectral information with separate encoders before fusing them. We evaluate DCCE on three coda-understanding tasks: social unit, coda type, and individual identification. We compare DCCE against strong baselines from literature under a common linear-probe protocol and analyze the role of each channel and training objective through ablations. On 1,383 codas, shared partitions and validation-only layer selection support three random splits and three recording-date-disjoint folds. DCCE with unit pairing and type supervision achieves mean random-split macro-F1 of 0.901/0.865/0.597 for unit/type/identity. Raw ICI remains strongest for type (0.987), and ICI plus temporal mel statistics is a strong identity reference (0.724). An auxiliary identity objective improves spectral-branch identity prediction but does not improve the mean fused-identity score in this configuration. Date-grouped evaluation and four year holdouts expose generalization limits. The results support task-dependent combinations of timing and spectral information, without establishing independent biological channels or general architectural superiority. Code, notebooks, and complete results are available at https://dcce-research-site.vercel.app.

cs.SD↗

arXiv2026

Experimental determination of the double $K$-shell vacancy creation probability in the electron-capture decay of $^{55}$Fe with a microcalorimeter

The probability of double $K$-shell vacancy production following electron-capture decay is determined for $^{55}$Fe using a high-resolution microcalorimeter. The $K$-hypersatellite lines arising from double $K$-shell-vacancy filling are directly resolved within the measured spectra of X-ray emissions which are dominated by Mn $K$-X-ray diagram lines. The energies of the Mn \textit{K}-hypersatellite lines, $Kα_2^h$, $Kα_1^h$, and $Kβ_{1,3}^h$, are measured as $(6143.5\pm0.5)$ eV, $(6162.1\pm1.5)$ eV and $(6811.9\pm2.6)$ eV, respectively. The relative intensities of \textit{K}-hypersatellite to diagram X-rays are measured to be $(1.30 \pm 0.07) \times 10^{-4}$. From the relative intensities together with established fluorescence yield values, the probability of double-vacancy creation per $K$-shell capture ($P_{KK}$) is determined as $(1.22 \pm 0.09) \times 10^{-4}$. Comparisons with previous experimental results and theoretical calculations are discussed.

nucl-ex↗

arXiv2026

Statistical Strong Lensing V. Inference with Microlensed Quasars

Microlensing of strongly lensed quasars provides direct information on the stellar component of lens galaxies, and can help break the degeneracy between the stellar initial mass function (IMF) and the dark matter. However, a statistical framework to forward model microlensing data while accounting for strong lensing selection effects is still missing. We incorporated microlensing effects into the hierarchical model of Paper III, and investigated to what extent flux measurements of strongly lensed quasars can break the degeneracy between the stellar IMF and the dark matter distribution. We tested the method on a mock sample of 219 doubly imaged quasars, lensed by circularly symmetric lenses. When microlensing effects are modelled, the inference using image fluxes recovers the input foreground population parameters and provides tighter constraints than the inference without flux constraints. The improvement is clearest in the $μ_α-μ_5$ plane, where $μ_α$ describes the mean stellar population synthesis mismatch parameter $\logα_{\rm SPS}$ and $μ_5$ describes the mean projected dark matter mass within $5\,{\rm kpc}$. The 68% credible interval on $μ_α$ shrinks from 0.16 dex to 0.09 dex, with the degeneracy between $μ_α$ and $μ_5$ significantly reduced. By contrast, fitting the image fluxes without accounting for microlensing leads to large biases on the inferred parameters. This work is a first step towards an accurate modelling of microlensing data. Application to real lens samples will require dropping the circular symmetry assumption and adopting more realistic models for the microlensing magnification.

astro-ph.GA↗

arXiv2026

A Finite Element Method for the Surface Biharmonic Problem without Stabilization or Piola Transforms

For the biharmonic problem on closed smooth surfaces, we construct a nonconforming finite element method that requires neither stabilization nor the Piola transform. The reduced enriched quartic (REQ) element relies solely on vertex values and scalar edge moments. The same local construction applies to every triangle of a piecewise planar surface. The bilinear form consists of elementwise Laplacian inner products and permits standard finite element assembly. By separating the planar vertex residual from the angle defect contribution, we establish weak discrete Miranda--Talenti-type estimates on both planar meshes and closed piecewise planar surfaces. These estimates guarantee uniform stability on sufficiently refined interpolating surface meshes. We prove the well-posedness of the method, first-order convergence in the energy norm, and second-order convergence in the broken $H^1$ and $L^2$ norms. Numerical tests on three surfaces show the predicted convergence rates and examine the discrete Miranda--Talenti constants and angle defect scaling.

math.NA↗

arXiv2026

NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness

Recursive self-improvement (RSI) requires a concrete mechanism through which an AI system observes its capabilities and converts that evidence into the next round of learning. We present NeoHorse-1, a family of agent-native models developed to explore this path through agentic post-training. Our system combines a heterogeneous model pool with intelligent routing, recording the predicted capability demand, selected service tier, and subsequent interaction for each user turn. These records are converted into training examples that preserve interleaved reasoning, tool calls, and harness context, and are admitted through structural validation, six-dimensional semantic evaluation, and subscene-level labeling. Routing signals organize supervised fine-tuning into a three-stage curriculum and extend to routing-guided on-policy distillation, where a teacher supervises student-generated responses under the same progression. Capability-guided allocation then converts evaluation feedback into the next training mixture, closing an evaluation-selection-update loop in which what the system learns to do shapes what it learns from next. Across eleven benchmarks covering harness-based agents, tool use, coding, and instruction following, post-training raises the macro-average from 58.94 to 64.87 at 4B and from 65.60 to 69.04 at 9B, substantially narrowing the aggregate gap between the post-trained 4B model and the 9B base model. We further adapt NeoHorse-1-4B into NeoHorse-Jev-4B, a prefill-only model that returns structured decisions and probabilities for agent workflows. NeoHorse-1 provides an initial prototype of this feedback-driven process and a path toward harness-mediated RSI across successive iterations.

cs.CL↗

arXiv2026

Spacecraft Pose Tracking via Kinematic Prediction and Selective Event Correction

In initialized spacecraft tracking, a short-horizon motion prediction remains available as event support changes. Observation-Aligned Event Innovation (OA-EI) uses events to correct selected components of this prediction. A constant-body-twist model propagates the complete pose; the event encoder proposes rotational and normalized image-plane translation corrections, retaining propagated axial depth. Candidate magnitudes come from a bounded Scale Gate and a deterministic activity factor, with empty windows returning the prediction unchanged. Evaluation uses four FRESH physical-event trajectories, twenty rollout starts, and three training seeds. Each rollout starts from two reference poses and then uses self-estimates. At 7.7 seconds, rotation and translation errors fall from 41.63°/57.28 mm for prediction alone to 15.73°/22.41 mm for OA-EI, reductions of 62.2% and 60.9%. Failure rates fall from 20.0% to 1.67%. Matched ablations favor the rotation-bearing interface for overall pose accuracy; combining scaling with activity support gives the lowest errors at every tested event-retention level. On Jetson Orin Nano, the measured in-memory update takes 44.56 milliseconds at the median and 58.77 milliseconds at P95.

cs.RO↗

arXiv2026

Trajectory Prediction for Autonomous Driving in Road Construction Zones Using Human-in-the-Loop Driving Data

Ego-trajectory prediction in urban roadway work zones remains an underexplored topic in autonomous-driving research, with limited dedicated data and systematic evaluation. Roadway work zones involve temporary changes in the lane layouts and connectivity of the roads, complicating interactions with surrounding vehicles and workers. These changes make future vehicle trajectories difficult to predict. To address these challenges, we collect WZDrive-Data, a multimodal dataset comprising 1,183 work-zone driving segments from 33 drivers, using a human-in-the-loop driving simulator with prerecorded worker animations and VR-controlled workers. Driver control and gaze signals are recorded from human participants, while vehicle trajectories and scene information are obtained from the simulator. Our proposed multimodal ego-trajectory predictor, WZDrive, achieves a mean minADE (K=6) of 1.09 m and minFDE (K=6) of 2.12 m. Compared with the best-performing state-of-the-art baseline for each metric in our benchmark, WZDrive reduces minADE (K=6) by 7.18% and minFDE (K=6) by 13.18%. This research addresses a gap in autonomous-driving research on urban work zones by combining human-in-the-loop data collection with multimodal learning from human driving signals and in the context of work zones.

cs.RO↗

arXiv2026

LongEmo: Towards Emotion Understanding and Reasoning in Long Videos

While recent Multimodal Large Language Models (MLLMs) have shown promise in affective computing, their reasoning capabilities are largely confined to short video clips with limited interactions. However, real-world emotions are not merely isolated instantaneous reactions but dynamic and cumulative processes deeply shaped by past experiences and ongoing events. To bridge this gap, we introduce LongEmoBench, a benchmark dedicated to emotion understanding and reasoning in long videos. It assesses progressive capabilities scaling from continuous scene interactions to complex episodic developments. Furthermore, we propose LongEmo, a novel memory-augmented agentic framework designed to tackle the immense challenges of long-range affective reasoning. LongEmo processes continuous video streams to construct an Event Memory Graph, explicitly modeling long-range dependencies and capturing emotional dynamics across discrete events. Given a question, the agent retrieves a query-relevant event stream from the graph, iteratively integrating multimodal memories and relational dependencies to deduce the final answer. Extensive evaluations of 17 representative methods reveal that they struggle significantly with emotion understanding and reasoning in long videos. In contrast, LongEmo achieves state-of-the-art performance, demonstrating the efficacy of its event-centric memory architecture.

cs.CV↗

arXiv2026

FloCA: Towards Faithful and Logically Consistent Flowchart Reasoning

Flowchart-oriented dialogue (FOD) systems aim to guide users through multi-turn decision-making or operational procedures by following a domain-specific flowchart to achieve a task goal. In this work, we formalize flowchart reasoning in FOD as grounding user input to flowchart nodes at each dialogue turn while ensuring node transition is consistent with the correct flowchart path. Despite recent advances of LLMs in task-oriented dialogue systems, adapting them to FOD still faces two limitations: (1) LLMs lack an explicit mechanism to represent and reason over flowchart topology, and (2) they are prone to hallucinations, leading to unfaithful flowchart reasoning. To address these limitations, we propose FloCA, a zero-shot flowchart-oriented conversational agent. FloCA uses an LLM for intent understanding and response generation, while delegating flowchart reasoning to an external tool that performs topology-constrained graph execution, ensuring faithful and logically consistent node transitions across dialogue turns. We further introduce an evaluation framework with an LLM-based user simulator and five new metrics covering reasoning accuracy and interaction efficiency. Extensive experiments on FLODIAL and PFDial datasets highlight the bottlenecks of existing LLM/VLM-based methods and demonstrate the superiority of FloCA. The code and dataset are publicly available at https://github.com/Jinzi-Zou/FloCA-flowchart-reasoning.

cs.AI↗

arXiv2026

Compression Beyond the Uncompressed: A Two-Stage Training Recipe for Soft Context Compression in RAG

Retrieval-Augmented Generation (RAG) improves knowledge-intensive generation by conditioning language models on retrieved documents, but processing these documents becomes increasingly expensive as retrieval depth grows. Soft context compression reduces this cost by encoding documents into compact continuous representations that can be precomputed and reused across queries. However, many existing methods train compressed models by distilling from a full-context teacher. When the teacher is wrong, such distillation can reinforce its errors, while teacher imitation provides no direct signal for improving beyond the teacher. We propose DEX-Comp, a two-stage training recipe that separates reliable imitation from targeted exploration. Pure Distillation learns only from teacher-correct questions to mitigate error propagation, while Hard Exploration applies outcome-based reinforcement learning to teacher-failed questions to directly optimize answer correctness. Across five open-domain QA benchmarks and retrieval depths from top-$5$ to top-$30$, DEX-Comp at $16\times$ compression outperforms all evaluated compression baselines and surpasses the untuned full-context RAG model in average accuracy, while reducing time-to-first-token by $4.4\times$--$23.7\times$. Evaluations across additional datasets and backbones further demonstrate its generalization.

cs.CL↗