arXiv subjects

Junwei Zhou

Publications and source records attributed to Junwei Zhou.

At least 19 recordsLinked to original sources

arXiv2026

Mockenhaupt's Three-Term Hardy-Littlewood Majorant Conjecture

For an integer $k \ge 0$, let $f_k(x) = 1 + e(x) + e((k+2)x)$ and $g_k(x) = 1 + e(x) - e((k+2)x)$ on $\mathbb{T} = \mathbb{R}/\mathbb{Z}$, where $e(x) = e^{2\pi i x}$. Mockenhaupt conjectured that $\|g_k\|_{L^p(\mathbb{T})} > \|f_k\|_{L^p(\mathbb{T})}$ whenever $2k < p < 2k+2$. The conjecture was previously known for $k \le 5$. We give a single analytic proof valid for every $k \ge 4$; in particular, this settles all previously open cases $k \ge 6$ and establishes the conjecture for every $k \ge 0$. The proof reduces the norm comparison to resonant Fourier coefficients on the two-torus and represents these coefficients, after analytic continuation, by triple-Bessel integrals. Neumann's product formula and the Weber-Schafheitlin formula yield a quantitative positive lower bound for the leading mode, while the remaining odd modes are controlled by a uniform tail estimate. The leading mode is then shown to dominate the tail for every $k \ge 4$. AI Usage. The mathematical argument of this paper was produced by the auto-research system Apex Math, an AI system built by Apex Intelligence. See Appendix A for the complete AI usage statement.

math.CA↗

arXiv2026

Contractibility of the complex of incompressible Seifert surfaces: the knot case of Kakimizu's problem

Let $K \subset S^3$ be a non-trivial knot and let $IS(K)$ be the simplicial complex whose vertices are the ambient isotopy classes of incompressible Seifert surfaces in the exterior $E(K)$, a finite set of distinct vertices spanning a simplex exactly when its classes admit simultaneously pairwise disjoint representatives. Kakimizu proved that $IS(K)$ is connected; whether it is contractible was asked by Przytycki and Schultens, who also identified the obstruction, namely that the projection used in the connectedness proof is not known to be well defined on isotopy classes. We prove that $IS(K)$ is contractible, and likewise every truncation $IS_\ell(K)$ on the vertices of genus at most $\ell$. The argument factors through a purely combinatorial theorem: every non-empty connected flag exchange complex is contractible, where an exchange complex carries a complexity function subject to two axioms which require only that an exchanging vertex exist, never that one be selected. That is what circumvents the obstruction. We locate the combinatorial theorem precisely against the existing literature -- it is not implied by dismantlability, and, when all descending links are finite, it follows from Zaremsky's descending-link criterion by a reindexing (Section 7.4) -- and we record what remains genuinely open. Two extensions are proved: the truncations above, and the case of links satisfying a linking condition that forces every spanning surface to be connected.

math.GT↗

arXiv2026

R$^2$A: Learning Persona Policies Through Persona Representation Learning and Runtime Alignment

The same Persona behavior can be beneficial in one context but harmful in another, causing static Persona elicitation to perform inconsistently across tasks. We introduce the Persona Selection--Realization Framework, which models behavior generation through a latent Persona state and decomposes it into Persona Selection and Persona Realization. The discrepancies between static Persona elicitation and an ideal Persona policy in these two components define the Selection Gap and Realization Gap, respectively. Building on this framework, we propose R$^2$A, a two-stage approach for learning Persona policies. Persona Representation Learning uses structured Who--How--What presentations to encode the target Persona's objective, conditional behavioral principles, and trajectory-level manifestations. Persona Runtime Alignment then removes the explicit Persona specification and jointly calibrates behavior selection and trajectory realization using task feedback. Across 12 evaluation settings covering the four principles of the Accountable-Professional Persona studied in this work, R$^2$A overall outperforms both the base model and static Persona elicitation. Ablation results further show that Persona Representation Learning is critical for preventing Runtime Alignment from producing behaviorally imbalanced policies and for achieving more stable Persona policy learning.

cs.CL↗

arXiv2026

ProRetrieval: Learning to Orchestrate Hybrid Search via Executable Program Synthesis

Real-world retrieval often composes structured constraints with semantic intents over text and images through arbitrary Boolean logic. Existing hybrid pipelines such as reciprocal rank fusion or self-querying retrievers admit only a fixed form of composition, while recent reinforcement-learning retrievers train the language model as a query generator for a single backend, leaving the orchestration of heterogeneous retrieval paths outside its action space. We propose ProRetrieval, which recasts the language model as a retrieval orchestrator: given a natural-language query, it synthesizes an executable program in a hybrid DSL interleaving SQL operators over structured fields with vector-retrieval primitives over text and images, with SQL itself providing the logical algebra that fuses heterogeneous candidate sets. We train Qwen3-4B with GRPO and DAPO under a hierarchical four-term reward, and evaluate on two new benchmarks built from Amazon products and Enron email. Our 4B model surpasses GPT-5.5 (Hit@1 0.81 vs. 0.69 on e-commerce; 0.91 vs. 0.86 on email) and Claude Opus 4.7 and a comprehensive suite of retrieval, LLM-augmented, structured-query, and graph-based baselines. Code: https://anonymous.4open.science/r/ProRetrieval/; data: https://huggingface.co/datasets/anonymous-7219/ProRetrieval.

cs.IR↗

arXiv2026

DocPC: Document-Level Visual Retrieval via Representative Page Composition

Visual document retrieval has advanced by encoding page screenshots with vision-language models, bypassing OCR pipelines. However, existing methods remain page-centric, misaligned with real-world scenarios requiring complete document retrieval. A naive page-then-document aggregation suffers from linear indexing cost and degraded retrieval when relevance spans multiple pages. We propose DocPC, a document-level visual retrieval framework based on Representative Page Composition: selecting representative pages and composing them into a single grid image for document-level indexing, reducing indexed images, vectors, and storage by 10.1x and end-to-end indexing time by roughly 7.7x. To handle multi-positive supervision prevalent at the document level, we combine multi-positive contrastive learning with sparsely scheduled listwise optimization. We also introduce DocViRe, a benchmark with multi-positive relevance annotations. DocPC-ColQwen achieves NDCG@5 of 44.09 on DocViRe, outperforming the strongest page-level baseline at 38.91 while reducing storage by 10.1x. Code is available at https://anonymous.4open.science/r/DocPC-Document-Level-Visual-Retrieval-via-Representative-Page-Composition-1D52. Data is available at https://huggingface.co/datasets/anonymous-7219/docpc.

cs.IR↗

arXiv2026

ASI-Bench: At the Dawn of Artificial Superintelligence

Artificial superintelligence (ASI) requires AI to move beyond mastering existing knowledge toward exploring the unknown, creating new knowledge, and turning new ideas into verifiable results. However, the capabilities of today's AI systems are still largely built on learning, compressing, and applying existing human knowledge. Accordingly, existing benchmarks primarily test whether AI can produce correct answers based on learned knowledge, or whether it can complete tasks under extensive human guidance. We therefore introduce ASI-Bench, the first benchmark to jointly evaluate AI systems' capabilities of innovative exploration and autonomous scientific execution across general research domains, and the first to progressively withdraw human methodological guidance within the same research project to test how far AI can proceed on its own. Built by over 40 experts with the cost of 31,000+ human hours, ASI-Bench contains 60 project-level research tasks across 11 scientific domains and progressively reduces methodological guidance to test whether AI can independently select methods, conduct research, and produce verifiable results. All tasks undergo expert review, AI-assisted auditing, sandbox execution, and scorer validation. Across 18 state-of-the-art agent--model configurations, the average score drops from 50.91 with full methodological guidance to 29.10 with only the method specified and 26.62 when agents must determine the method themselves. This sharp decline shows that current systems remain heavily dependent on human guidance and are still far from autonomously conducting end-to-end, project-level scientific research. ASI-Bench is open to the world. We invite researchers and builders everywhere to contribute new tasks, challenge the limits of today's AI, and help accelerate humanity's collective path toward artificial superintelligence at https://asibench.apexin.ai/submit.

cs.AI↗

arXiv2026

Focal-point scanning for dose delivery and optimization with focused laser-accelerated very-high-energy electron beams

Focused very-high-energy electron (VHEE) beams can produce localized dose enhancement at selected depths, but irradiation of a finite target requires coordinated control of multiple focal positions, incidence directions, and beam weights while limiting exposure of nearby organs at risk (OARs). We present Focal-Point Scanning (FPS), a dose delivery and optimization method developed for laser wakefield accelerator (LWFA)-driven VHEE beams. The method is based on a two-dipole focusing system that produces single-plane beam convergence and allows the focal position to be varied by changing the magnetic field strength. FPS distributes focal points throughout the planning target volume and determines focal-point-specific incidence sectors according to the geometry of nearby critical OARs. The method was evaluated using the AAPM TG119 C-shape benchmark and one previously treated lung radiotherapy case. At matched target coverage, FPS reduced the TG119 Core mean dose by approximately one half relative to parallel VHEE and intensity-modulated x-ray plans, approaching the single-field proton pencil-beam-scanning reference. In the lung case, FPS maintained target coverage comparable to the clinical volumetric modulated arc therapy reference while reducing the mean dose to every evaluated OAR; spinal-cord mean and maximum doses decreased by 93.2% and 87.2%, respectively. The evaluated OAR mean doses varied little across rms energy spreads of 0 to 10% and for a flat-top electron spectrum spanning 150 to 250 MeV. These results demonstrate that focal-point-specific angular selection can translate focused-beam physics into effective OAR sparing and support FPS as a planning strategy for broadband LWFA-VHEE radiotherapy.

physics.med-ph↗

arXiv2026

IDEAL-Bench: Indoor Dataset and Evaluation suite for Analyzing 3D Layout reasoning

Spatial question answering is the dominant paradigm for evaluating spatial intelligence in Vision-Language Models (VLMs), but it leaves a complementary axis of spatial competence under-evaluated: holistic 3D layout inference, which predicts every visible object's pose and extent from a single image in a structured form. To this end, we introduce IDEAL-Bench, an evaluation suite that requires VLMs to predict structured 3D layouts on photorealistic indoor scenes across 10 room types, scored along five numerical dimensions and a perceptual render-and-compare protocol. By operating on semantically realistic scenes with full asset substitution under controlled lighting and viewpoint, IDEAL-Bench moves beyond CLEVR-style simple geometric primitives so that any image-space discrepancy reflects spatial reasoning alone. The benchmark is built on IDEAL-Scenes, a procedurally generated dataset of 1,000 re-renderable Blender environments with ground-truth layouts. Evaluating 15 prominent VLMs reveals three findings: the task remains substantially unsolved, with the strongest model reaching only 62.1/100 overall; all models exhibit a sharp asymmetry between object recognition and geometric regression, indicating that current VLMs are trained to describe scenes rather than to measure them; model rankings partially diverge from those on QA-based and primitive-reconstruction benchmarks: top-tier consensus holds, but mid-tier rankings shift substantially. Collectively, these findings establish IDEAL-Bench as a diagnostic suite, targeting the geometric and structural competencies that QA-based evaluation cannot surface, and paving the way towards more rigorous evaluation of spatial intelligence in next-generation VLMs. Together, these findings position IDEAL-Bench as a principled diagnostic for whether future VLMs achieve genuine spatial understanding rather than linguistic approximations of it.

cs.CV↗

arXiv2026

Perceive-then-Plan: Layout-as-Policy for Monocular 3D Scene Layout Estimation

Building structured 3D scene layouts from a single image requires reconciling visual observations with physical and spatial constraints, a challenge that is difficult to address with direct prediction alone. In this work, we formulate monocular 3D layout estimation as a perceive-then-plan problem with vision-language models, where a Perceiver first grounds the 3D objects and then a Planner iteratively refines the scene hypothesis through actions that improve physical plausibility while preserving consistency with the input image. We propose Layout-as-Policy (LaP), which casts the planning stage as a policy learning problem: 3D layouts are represented as structured states, and refined via discrete actions such as translation, rotation, and rescaling. Starting from an observation-aligned initialization with the geometry-enhanced Perceiver, the LaP Planner is trained to produce action sequences that progressively resolve geometric inconsistencies and enforce realistic spatial relations. To enable effective learning, we combine supervised trajectory initialization with preference-based optimization, allowing the model to learn corrective behaviors without requiring explicit reward engineering. This formulation transforms layout estimation from a one-shot prediction task into an iterative refinement process, enabling better handling of global constraints and complex object interactions. Experiments demonstrate that our approach produces layouts that are more physically coherent and better aligned with visual observations, while naturally supporting downstream tasks such as scene editing and manipulation.

cs.CV↗

arXiv2026

Distributed Image Compression with Multimodal Side Information at Extremely Low Bitrates

Distributed Image Compression (DIC) is crucial for multi-view transmission, especially when operating at extremely low bitrates (< 0.1 bpp). Its core challenge is effectively utilizing side information to achieve high-quality reconstruction under strict bitrate budgets. However, existing DIC approaches struggle to exploit global context and object-level details from side information, leading to local blurring and the loss of fine details in the reconstruction. To address these limitations, we propose a Multimodal DIC framework (MDIC), which, for the first time, leverages side information in a multimodal manner into the DIC paradigm, effectively preserving fine-grained local details and enhancing global perceptual quality in reconstructed images. Specifically, we introduce a text-to-image diffusion-based decoder conditioned on textual side information extracted from correlated images to capture shared global semantics. Moreover, we design a feature-mask generator, supervised by a multimodal fine-grained alignment task, to strengthen the exploitation of visual side information. The generated mask serves two purposes: first, it guides the extraction of fine-grained details from losslessly transmitted side information to preserve the semantic consistency of reconstructed details; second, it regulates the extraction of clustered feature representations from the quantized VQ-VAE embeddings, compensating for category information lost under the extreme compression of the primary image. Extensive experiments on the widely used KITTI Stereo and Cityscapes datasets demonstrate that MDIC achieves state-of-the-art perceptual quality at extremely low bitrates.

cs.CV↗

arXiv2026

SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks

Agent Skills are structured packages of procedural knowledge that augment large language model (LLM) agents at inference time. Despite rapid adoption, there is no standard way to measure whether they actually help. We present SkillsBench, a benchmark whose current inventory contains 87 tasks across 8 domains paired with curated Skills and deterministic verifiers. Our latest aggregate evaluation runs the 87-task benchmark under matched no-Skills and curated-Skills conditions for 18 model-harness configurations. Curated Skills raise the average pass rate from 33.9% to 50.5% (+16.6 percentage points; 25.5% normalized gain), with configuration-level gains ranging from +4.1 to +25.7 pp. Focused Skills with at most three modules outperform larger or exhaustive bundles, and smaller models with Skills can match larger models without them. SkillsBench establishes paired evaluation as the foundation for rigorous measurement of Skill efficacy on agentic, expertise-heavy work.

cs.AI↗

arXiv2026

An Impulse-formed Navier-Stokes Solver based on Long-range Particle Flow Maps

We present a particle-grid characteristic-mapping framework that extends long-range characteristic mapping from inviscid flows to general Navier-Stokes dynamics with viscosity, body forces, and complex boundaries. Unlike traditional grid-based and vorticity-centered characteristic methods, our method is built on the observation that particle trajectories naturally provide the long-range flow map, enabling geometric quantities and their gradients to be transported in a direct and effective manner. We identify the impulse, the gauge variable of the velocity field, as the primary quantity mapped along characteristics while remaining compatible with standard velocity-based incompressible solvers. Using the 1-form representation of the impulse equation, we derive an integral formulation that decomposes the impulse evolution into a component transported geometrically along the particle flow map and a complementary component generated by viscosity and body forces evaluated through path integrals accumulated along particle trajectories. These components together yield a unified characteristic-mapping solver capable of handling incompressible Navier-Stokes flows with viscosity and body forces while maintaining the accuracy and geometric fidelity of characteristic transport.

cs.CE↗

arXiv2026

A Level Set Method on Particle Flow Maps

This paper introduces a Particle Flow Map Level Set (PFM-LS) method for high-fidelity interface tracking. We store level-set values, gradients, and Hessians on particles concentrated in a narrow band around the interface, advecting them via bidirectional flow maps while using a conventional grid-based representation elsewhere. By interpreting the level set value as a 3-form and its gradient as a 1-form, PFM-LS achieves exceptional geometric fidelity during complex deformations and preserves sub-grid features that traditional methods cannot capture. Our dual-timescale approach utilizes long-range maps for values and gradients, with frequent reinitialization of short-range maps for the distortion-sensitive Hessian, alongside adaptive particle control that maintains sufficient density within the narrow band. We also develop a hybrid particle-grid quasi-Newton redistancing scheme that preserves fine-scale features while enforcing the signed-distance property. Benchmark comparisons in 2D and 3D demonstrate that PFM-LS achieves state-of-the-art volume preservation and shape fidelity against a broad range of existing level-set methods.

physics.comp-ph↗

arXiv2026

DSL-R1: From SQL to DSL for Training Retrieval Agents across Structured and Unstructured Data with Reinforcement Learning

Effective retrieval in complex domains requires bridging the gap between structured metadata and unstructured content. Existing systems typically isolate these capabilities, relying on either symbolic filtering or vector similarity, failing to capture their interplay. In this work, we propose DSL-R1, a unified framework that synergizes logical reasoning with semantic matching via a novel Domain-Specific Language (DSL). By embedding vector primitives within SQL-style operators, our approach leverages the complementary strengths of symbolic precision and semantic coverage. We further introduce a reinforcement learning mechanism where rule-based execution feedback and retrieval quality rewards jointly optimize the DSL generation, balancing structural correctness and semantic alignment. Evaluations on a large-scale industrial email benchmark demonstrate that DSL-R1 achieves a +12.3% improvement in Hit@1/3, consistently outperforming decoupled baselines and establishing a robust paradigm for hybrid retrieval.

cs.IR↗

arXiv2025

GENA3D: Generative Amodal 3D Modeling by Bridging 2D Priors and 3D Coherence

Generating complete 3D objects under partial occlusions (i.e., amodal scenarios) is a practically important yet challenging problem, as large portions of object geometry are unobserved in real-world scenarios. Existing approaches either operate directly in 3D, which ensures geometric consistency but often lacks generative expressiveness, or rely on 2D amodal completion, which provides strong appearance priors but does not guarantee reliable 3D structure. This raises a key question: how can we achieve both generative plausibility and geometric coherence in amodal 3D modeling? To answer this question, we introduce GENA3D (GENarative Amodal 3D), a framework that integrates learned 2D generative priors with explicit 3D geometric reasoning within a conditional 3D generation paradigm. The 2D priors enable the model to plausibly infer diverse occluded content, while the 3D representation enforces multi-view consistency and spatial validity. Our design incorporates a novel View-Wise Cross-Attention for multi-view alignment and a Stereo-Conditioned Cross-Attention to anchor generative predictions in 3D relationships. By combining generative imagination with structural constraints, GENA3D generates complete and coherent 3D objects from limited observations without sacrificing geometric fidelity. Experiments demonstrate that our method outperforms existing approaches in both synthetic and real-world amodal scenarios, highlighting the effectiveness of bridging 2D priors and 3D coherence in generating plausible and geometrically consistent 3D structures in complex environments.

cs.CV↗

arXiv2025

CoCo4D: Comprehensive and Complex 4D Scene Generation

Existing 4D synthesis methods primarily focus on object-level generation or dynamic scene synthesis with limited novel views, restricting their ability to generate multi-view consistent and immersive dynamic 4D scenes. To address these constraints, we propose a framework (dubbed as CoCo4D) for generating detailed dynamic 4D scenes from text prompts, with the option to include images. Our method leverages the crucial observation that articulated motion typically characterizes foreground objects, whereas background alterations are less pronounced. Consequently, CoCo4D divides 4D scene synthesis into two responsibilities: modeling the dynamic foreground and creating the evolving background, both directed by a reference motion sequence. Given a text prompt and an optional reference image, CoCo4D first generates an initial motion sequence utilizing video diffusion models. This motion sequence then guides the synthesis of both the dynamic foreground object and the background using a novel progressive outpainting scheme. To ensure seamless integration of the moving foreground object within the dynamic background, CoCo4D optimizes a parametric trajectory for the foreground, resulting in realistic and coherent blending. Extensive experiments show that CoCo4D achieves comparable or superior performance in 4D scene generation compared to existing methods, demonstrating its effectiveness and efficiency. More results are presented on our website https://colezwhy.github.io/coco4d/.

cs.CV↗

arXiv2025

Fluid Simulation on Vortex Particle Flow Maps

We propose the Vortex Particle Flow Map (VPFM) method to simulate incompressible flow with complex vortical evolution in the presence of dynamic solid boundaries. The core insight of our approach is that vorticity is an ideal quantity for evolution on particle flow maps, enabling significantly longer flow map distances compared to other fluid quantities like velocity or impulse. To achieve this goal, we developed a hybrid Eulerian-Lagrangian representation that evolves vorticity and flow map quantities on vortex particles, while reconstructing velocity on a background grid. The method integrates three key components: (1) a vorticity-based particle flow map framework, (2) an accurate Hessian evolution scheme on particles, and (3) a solid boundary treatment for no-through and no-slip conditions in VPFM. These components collectively allow a substantially longer flow map length (3-12 times longer) than the state-of-the-art, enhancing vorticity preservation over extended spatiotemporal domains. We validated the performance of VPFM through diverse simulations, demonstrating its effectiveness in capturing complex vortex dynamics and turbulence phenomena.

cs.GR↗

arXiv2025

R1-Code-Interpreter: LLMs Reason with Code via Supervised and Multi-stage Reinforcement Learning

Practical guidance on training Large Language Models (LLMs) to leverage Code Interpreter across diverse tasks remains lacking. We present R1-Code-Interpreter, an extension of a text-only LLM trained via multi-turn supervised fine-tuning (SFT) and reinforcement learning (RL) to autonomously generate multiple code queries during step-by-step reasoning. Unlike prior RL + tool-use efforts focused on narrow domains such as math or retrieval, we curate 144 diverse reasoning and planning tasks and show that training a general-purpose Code Interpreter across them presents significant challenges due to task heterogeneity and scarcity of effective samples. To address this, we introduce a multi-stage curriculum learning approach that partitions training samples by measured improvement potential. The RL training prioritizes samples with higher potential and gradually shifts to lower-potential ones, increasing the average RL gains from merely +3.4% to +9.3% across Qwen-2.5 models (3/7/14B). Our final model, R1-CI-14B, improves average accuracy on the 37 test tasks from 44.1% to 72.4%, outperforming text-only GPT-4o (58.6%) and GPT-4o with Code Interpreter (70.9%). Notably, R1-CI-14B also exhibits emergent self-checking behavior through code generation. Datasets, Codes, and Models are available at https://github.com/yongchao98/R1-Code-Interpreter and https://huggingface.co/yongchao98.

cs.AI↗