arXiv subjects

Jian Wang

Publications and source records attributed to Jian Wang.

At least 91 records · Page 5Linked to original sources

arXiv2026

Examples beyond Bounded Mean Motion for Quantitative Rigidity on the Two-Torus

This note supplies genuinely non-fibred examples for the manuscript: Rigidity on the Two-Torus and Sarnak's Conjecture. For every $0<δ<\tfrac12$, we construct $C^\infty$ Lebesgue-area-preserving pseudo-rotations of $\mathbb{T}^2$ which satisfy the $(C,δ)$-deviation condition but do not have bounded mean motion. We give both semi-irrational and totally irrational rotation vectors and two realizations: a controlled weakly mixing Anosov--Katok construction and an explicit weakly mixing special flow construction. Weak mixing is used as a conjugacy-invariant obstruction to every continuous circle-rotation factor. Consequently, none of the resulting maps is topologically conjugate, by a linear or nonlinear change of coordinates, to a skew product over a circle rotation. In the special-flow realization, the same lacunary Fourier series simultaneously gives weak mixing, the sharp upper bound $O(n^δ)$, and unbounded deviations; in fact no smaller deviation exponent is possible. The semi-irrational examples meet the assumptions of Theorems~1 and~2 of the cited manuscript, whereas the totally irrational examples meet those of Theorem~1. Each construction produces continuum many maps and continuum many topological conjugacy classes of each rotation type.

math.DS↗

arXiv2026

Scalar curvature on Kähler blow-ups and systolic inequalities

In this paper, we develop the weighted level set method for a Kähler manifold $(X^n,ω)$ admitting an almost holomorphic map to a possibly singular base $Z$, which is not uniruled. As a key intermediate result, we prove that any blowup $\operatorname{Bl}_SX$ of $X$ along smooth submanifolds $S$ of $ \operatorname{codim} S\ge2$ admits a sequence of Kähler metrics with scalar curvature globally and arbitrarily $C^0$-close to the scalar curvature of $ω$. As a consequence, we establish the sharp $2$-systole estimate for every positive scalar curvature Kähler manifold $(X,ω)$ and prove $\min_XS(ω) \cdot\operatorname{sys}_2(ω) \le 2πr(r+1)$, where $r$ is the rational dimension of $X$, with equality if and only if the universal cover splits as $(\widetilde X,\widetilde ω) \cong (\mathbb P^r,ω_{\mathrm{FS}}) \times(Y^{n-r},ω_{\mathrm{RF}})$ up to normalization where $ω_{\mathrm{FS}}$ is the Fubini-Study metric and $ω_{\mathrm{RF}}$ is Ricci-flat. We also show a sharp even-systolic inequality in the same setting when the general fibre is the projective space.

math.DG↗

arXiv2026

RecGPT-Mobile-V2 Technical Report

Personalized Query prediction maps implicit behavioral signals---clicks, favorites, purchases, and post-purchase exploration---to explicit retrieval intent. On-device deployment makes this task particularly challenging: behavioral trajectories are noisy and multi-scale, multiple Queries may be valid for a single trajectory, and a uniform reasoning policy either expends unnecessary computation on simple instances or allocates insufficient capacity to complex ones. We introduce RecGPT-Mobile-V2, an end-to-end framework that treats intent quality and execution efficiency as coupled objectives within a staged design. The framework transforms heterogeneous interactions into an evidence-preserving trajectory, establishes a recommendation-native foundation through domain adaptation and supervised alignment, and applies reasoning-cost optimization only after grouped rollouts meet grounding and utility criteria. The resulting teacher is distilled into a compact student deployed with low-bit execution, structured compression, and budget-aware device--cloud routing. In an aligned CoT ablation, an evidence-focused short rationale increases ROUGE-L from 0.228 to 0.315 and Jaccard from 0.174 to 0.248, while slightly outperforming the full five-stage rationale. In the controlled RL comparison, the complete reward formulation improves Query quality from 73.2% under quality-only RL to 78.6%, lowers the hard-failure rate from 3.6% to 1.6%, and reduces the median CoT length from 62 to 14 tokens. Online retrieval analysis further indicates that the Query recall channel retrieves inventory complementary to that surfaced by established recall channels. Collectively, these findings support sufficiency-oriented rather than uniformly short reasoning: retain decision-relevant evidence and allocate additional computation only when it is likely to improve the predicted Query.

cs.IR↗

arXiv2026

Stability for Helly-type and triangle-free families

We consider $k$-graphs, $\mathcal{F}\subset \binom{[n]}{k}$, $k\geq 3$. A $k$-graph is called intersecting if any two of its edges have non-empty intersection. It is called a star if all its edges share a common vertex. The $k$-graph $\mathcal{F}$ is called Helly if all its intersecting subfamilies are stars. If the same is required only for subfamilies consisting of three edges, it is called triangle-free. It is well known that for $n\geq 3k/2$, the full star is the unique largest triangle-free family whence the largest Helly family as well. In 1984 Tuza proved the best possible bound $|\mathcal{F}|\leq \binom{n-k-1}{k-1}+\binom{n-2}{k-2}+1$ for Helly families that are not stars, albeit only for some unspecified $n>n_0(k)$. The aim of this paper is twofold. First we establish the same bound for $n>2k$. Second we show that for $n>12k^2$ the same upper bound holds for triangle-free families. It is shown as well that it is not true for $2k<n\leq 3k-4$.

math.CO↗

arXiv2026

Retrieve, Match, Escalate: Accurate and Scalable Product Linking with VLM-Distilled Cross-Encoders and Agentic VLMs

Product linking, the entity-resolution task of mapping merchant product records to canonical catalog products, consolidates fragmented listings so downstream search, recommendation, and advertising see one clean entry per product. At marketplace scale, billions of noisy, multi-category records must be resolved against tens of millions of canonical products, where scoring every candidate with a single model is either too weak for the hard cases or too costly for the easy ones. We present a production retrieve-then-match cascade that spends computation in proportion to difficulty: retrieval surfaces plausible matches, a lightweight text cross-encoder auto-resolves the high-confidence majority, and an agentic multimodal vision-language model settles the ambiguous remainder by inspecting product images and issuing web searches for evidence that is in neither record. The cross-encoder is distilled from millions of dual-VLM-consensus labels, retiring human annotation from the training set, and is calibrated to auto-accept links at a 98% precision bar validated against a smaller operator-certified audit. The agent is a self-hosted open-weight model that reaches a closed frontier VLM's precision at a four-point recall cost (88% versus 92%) for roughly one-seventh the per-pair cost, with no fine-tuning. Per-pair cost spans nearly five orders of magnitude from the cheap cross-encoder to the frontier VLM, so escalating only the hard tail to the agent raises end-to-end link coverage from the cheap stage's 68% to 77%.

cs.AI↗

arXiv2026

SEAM: Shot Entity-Attribute Memory for Consistent Short-Drama Generation at Scale

Short-drama generation has grown into a large, industrialized pipeline, and as it scales from isolated shots to the episode level, visual continuity has become a critical bottleneck. Current agent frameworks generate each shot in isolation, so context drifts across shots and props, character posture, and blocking turn inconsistent. Once assembled, these small discrepancies amplify into severe visual breaks. We present SEAM (Shot Entity-Attribute Memory), a training-free, model-agnostic memory graph that repairs continuity entirely at the prompt-text layer by extracting a multi-dimensional state for every shot, retrieving only causally prior context over the resulting graph, filtering it selectively, and injecting the surviving constraints by natural-language prompt rewriting. We further release SEAM-Bench, a double-blind continuity storyboarding benchmark, on which SEAM raises cross-episode continuity recall from 0.700 to 0.946, generalizes across six mainstream text models, and yields consistent, though not yet significant, gains at the generated-image layer. Deployed as a mandatory stage in CreativeFitting's SEAM-Agent production pipeline over 201 shots, SEAM reaches a 96.5% director-acceptance rate with zero unsafe injections; a conservative counterfactual attributes at least 21.9 percentage points of that rate to its cross-episode memory.

cs.AI↗

arXiv2026

Sparse Principal Component Analysis with Energy Profile Dependent Sample Complexity

We study sparse principal component analysis in the high-dimensional, sample-limited regime, aiming to recover a leading component supported on a few coordinates. Despite extensive progress, most methods and analyses are tailored to the flat-spike case, offering little guidance when spike energy is unevenly distributed across the support. Motivated by this, we propose Spectral Energy Pursuit (SEP), an effective iterative scheme that repeatedly screens and reselects coordinates, with a sample complexity that adapts to the energy profile. We develop our framework around a structure function $s(p)$ that quantifies how spike energy accumulates over its top $p$ entries. To our knowledge, SEP is the first polynomial-time SPCA method with a sample-complexity guarantee governed by the full energy profile: it succeeds with $m\gtrsim \max_{1\le p\le k} p\,s^2(p)\,\log n$ samples, recovering the classical $k^2\log n$ rate for flat spikes and improving to $k\log n$ for sufficiently concentrated profiles. As a lightweight post-processing, a single truncated power iteration is proven to enable the final estimator to attain a uniform statistical error bound. Empirical simulations using a flat profile and offset-regularized decaying profiles validate that SEP adapts to profile structure without profile-specific tuning and outperforms existing algorithms.

cs.IT↗

arXiv2026

Know2Guess: A Contamination-Aware Multi-Zone Benchmark for Knowledge-Boundary Evaluation in Large Language Models

Reliable evaluation of large language models should separate supported answering from unsupported guessing without conflating either with data contamination, prompt idiosyncrasy, or generic refusal behavior. We present a contamination-aware, multi-zone benchmark for measuring the transition from answerable knowledge to abstention-expected unknowns under frozen build-time labels. The benchmark contains 1,200 items across five domains, explicit abstention expectations, contamination-risk metadata, and dual parsing with an official strict parser plus a normalized robustness parser. We evaluate FLAN-T5, Qwen2.5-Instruct, and Llama-3-Instruct models under locked answer-or-abstain prompts, answer-only controls, and prompt-template variants. The benchmark is not solved by generic non-answer behavior: FLAN baselines remain weak on productive abstention, while stronger instruction-tuned models expose a selective but incomplete transition from answering to abstaining. Qwen2.5-3B-Instruct achieves the best overall reliability, but answer-expected zones remain difficult, calibration remains poor, and benign-item refusal persists. Prompt and parser robustness analyses preserve the main ranking and qualitative conclusions. The benchmark therefore provides a reproducible protocol for auditing answerability, abstention, refusal, and contamination as distinct but interacting dimensions of LLM reliability.The dataset is publicly available at https://github.com/renweimeng/Know2Guess-A-Contamination-Aware-Multi-Zone-Benchmark.

cs.CL↗

arXiv2026

SAGE: Self-Evolving Storyboard Skills via Attribution-Guided Rule Evolution

Storyboards turn screenplays into visual shot plans for automated short drama production. Professional storyboarding relies on tacit directorial expertise and remains an industrial bottleneck. Large language models can automate this step, but methods for supplying directing knowledge face three challenges: (1) Knowledge acquisition: the craft remains implicit in exemplars or must be written manually. (2) Knowledge refinement: authored knowledge is not evaluated against execution outcomes, and opaque generation prevents feedback attribution to the knowledge behind each decision. (3) Knowledge injection: injecting all knowledge exceeds usable context, while manual selection for every narrative group does not scale. We present SAGE (Skill with Attribution-Guided Evolution), a deployed framework that learns, attributes, evolves, and routes directing knowledge from expert demonstrations. SAGE derives rules that are independent of episode content by contrasting each training screenplay with its expert storyboard. During generation, the model records each narrative group's adopted rules. Combining these records with localized feedback enables targeted updates to individual rules. Evolved rules form scenario packages with a routing index, so each group retrieves only a bounded set appropriate to its situation without expert intervention. On 18 test episodes across three genres, SAGE scored 77.8 on a rubric validated by experts, versus 77.1 for professional directors. Deployed for 14 days on Virtual Film Studio, SAGE produced 1,344 narrative group outputs; 87.2 percent were accepted without substantive edits, and the production team recorded over 83 percent less authoring time per episode. We release PROSE, the first public dataset pairing screenplays with storyboards by professional directors across 68 episodes: https://github.com/creDreams/PROSE.

cs.AI↗

arXiv2026

Beyond Control Points: Arcsecond Relative-Motion Estimation of Vision Measurement Platforms With Incomplete or Absent Control Fields

Long-range vision-based deformation monitoring is highly sensitive to motion of the camera platform. Absolute-pose differencing typically relies on dedicated control data and propagates two independent pose errors into the relative-motion estimate. We develop a control-adaptive differential framework that estimates inter-frame platform motion directly from image displacements and known 3D points. With no dedicated control point, the framework recovers platform rotation from measurement-point observations. One surveyed control point enables prior-constrained translation recovery, while two nonparallel control rays recover full 3D translation. The framework requires neither nonlinear optimization nor an initial pose estimate. Excluding control data from the rotation stage makes the rotation estimate exactly immune to contamination confined to the control field. The inherited differential formulation also cancels translational extrinsic errors exactly. We derive the rotation observability condition, a leakage bound for unmodeled translation and nonrigid point motion, and the single-point axial-prior bias law. Under 0.5-pixel image noise, attitude changes of up to 30~arcmin, and 3D point perturbations of up to 2~mm, the multi-camera estimator achieves a rotation RMSE of 2.97~arcsec and an average runtime of 0.46~ms. With one surveyed control point, its prior-constrained translation RMSE is 1.19~mm. In a bridge experiment without a stable control field, the median coordinate-wise displacement RMSE relative to total-station measurements is 0.85~mm. The estimator also maintains zero divergence under the tested 3D coordinate perturbations on public RGB-D and stereo sequences. These results establish state-of-the-art accuracy, calibration robustness, and computational efficiency among the evaluated methods.

cs.CV↗

arXiv2026

DREAM Technical Report

Industrial recommender systems commonly use cascaded retrieval, ranking, and re-ranking pipelines. Although efficient, these pipelines fragment information and objectives across modules, rely on rigid rules, and have limited awareness of real-time intent, leaving session-level shifts among browsing, comparison, and purchase insufficiently addressed. We present DREAM (Developing Recommender Engine with Agentic Methods), an autonomous optimization control architecture that adds a perception-aware, orchestrable, and auditable policy layer atop existing pipelines without replacing them. DREAM has two core components. First, a three-tier Intent Engine fuses on-device signals into structured L0/L1/L2 intent representations; its edge-cloud trigger chain reduces reporting volume to approximately 8.7%. Second, a Meta Engine uses a MetaModel for layered M1-to-M2-to-M3 reasoning: intent summarization, strategy planning informed by Strategy Memory, and parameter translation. It dispatches the resulting parameters through a unified outlet with safety guardrails. A Reward Dual Loop continuously optimizes both components by combining offline simulation for strategy-space exploration with online feedback for outcome calibration, forming a cycle of generation, execution, evaluation, and experience accumulation. Large-scale A/B tests on Taobao's homepage feed show that re-ranking control alone improves IPV by 2.06%, Core IPV by 2.39%, and GMV by 0.88%. Extending control to fine ranking raises these gains to 2.71%, 3.06%, and 1.31%, respectively, while consistently improving PV by more than 1%. These gains require neither replacement of pipeline models nor compromise of serving stability, supporting agentic meta-control as a viable paradigm for industrial recommendation.

cs.IR↗

arXiv2026

Dissipationless Photovoltaic Spin Hall Effect from Spin-current Vorticity

Spin-current vorticity (SCV) can generate the linear magnetic spin Hall effect [\href{https://doi.org/10.1038/s41586-018-0853-0}{Nature \textbf{565}, 627 (2019)}], yet its role in nonlinear spin Hall transport has been much less explored. Here, we show that, under a dc electric field, SCV can deflect optically excited electrons to drive a dissipationless photovoltaic spin Hall effect (PSHE), in which optical excitation by circularly and linearly polarized light is governed by the Berry curvature and quantum metric, respectively. Because the Berry curvature is $\mathcal{T}$-odd whereas the quantum metric is $\mathcal{T}$-even, their respective combinations with the $\mathcal{T}$-odd SCV give rise to $\mathcal{T}$-even and $\mathcal{T}$-odd PSHEs, where $\mathcal{T}$ denotes time-reversal symmetry. Remarkably, we find that the spin current of the $\mathcal{T}$-even PSHE can be reversed by switching the light helicity, as illustrated in monolayer WTe$_2$. By contrast, the $\mathcal{T}$-odd PSHE in altermagnets changes sign upon Néel-vector reversal, as demonstrated in a $d$-wave altermagnetic model. Beyond the PSHE, we show that the SCV dipole governs both the Drude and intrinsic nonlinear spin Hall effects proposed recently. Our results reveal two switchable spin Hall mechanisms and establish SCV as a unifying concept for understanding dissipationless nonlinear spin Hall transport.

cond-mat.mtrl-sci↗

arXiv2026

Nonembeddable contractible open manifolds arising from Whitehead doubling

We study a family of genus-one contractible open manifolds constructed by iterated Whitehead doubling from a nontrivial knot $K$ and an even half-twist $m$. For each such pair, we prove that the resulting contractible open $3$-manifold $W(K,m)$ does not embed as an open subset of any compact, locally connected and locally $1$-connected metric $3$-space. We also classify this family: $W(K,m)$ is homeomorphic to $W(K',m')$ if and only if $m=m'$ and $K$ is isotopic to $K'$. Thus the knot type and the half-twist parameter form a complete invariant for these nonembeddable contractible open manifolds. The proof combines the topology of ends with JSJ decompositions, hyperbolic pieces, and a rank estimate for iterated Whitehead doubled knot groups. The same method gives infinitely many pairwise non-homeomorphic higher-dimensional examples which embed in no compact, locally connected and locally $1$-connected metric space of the same dimension.

math.GT↗

arXiv2026

GAM-Agent: Game-Theoretic and Uncertainty-Aware Collaboration for Complex Visual Reasoning

We propose GAM-Agent, a game-theoretic multi-agent framework for enhancing vision-language reasoning. Unlike prior single-agent or monolithic models, GAM-Agent formulates the reasoning process as a non-zero-sum game between base agents--each specializing in visual perception subtasks--and a critical agent that verifies logic consistency and factual correctness. Agents communicate via structured claims, evidence, and uncertainty estimates. The framework introduces an uncertainty-aware controller to dynamically adjust agent collaboration, triggering multi-round debates when disagreement or ambiguity is detected. This process yields more robust and interpretable predictions. Experiments on four challenging benchmarks--MMMU, MMBench, MVBench, and V*Bench--demonstrate that GAM-Agent significantly improves performance across various VLM backbones. Notably, GAM-Agent boosts the accuracy of small-to-mid scale models (e.g., Qwen2.5-VL-7B, InternVL3-14B) by 5--6\%, and still enhances strong models like GPT-4o by up to 2--3\%. Our approach is modular, scalable, and generalizable, offering a path toward reliable and explainable multi-agent multimodal reasoning.

cs.AI↗

arXiv2026

Hybrid Token Compression for Vision-Language Models

Vision-language models (VLMs) rely on hundreds of visual tokens, leading to high computational and memory costs. Existing compression methods face a trade-off: continuous compression can weaken high-level semantics, while discrete quantization may lose fine-grained appearance details. We introduce HTC-VLM, a hybrid visual token compression framework that disentangles semantics and appearance through two complementary pathways. A continuous pathway preserves detailed ViT patch features, while a discrete pathway provides semantic anchors using MGVQ quantization represented by four tokens. The two pathways are fused into a 580-token hybrid sequence and compressed into a single token using a disentanglement attention mask and a bottleneck. Under the same one-token output budget, HTC-VLM achieves 87.2% average performance retention across seven benchmarks (GQA, VQAv2, MMBench, MME, POPE, SEED-Bench, and ScienceQA-Image), outperforming the leading continuous baseline at 81.0%. Attention analysis shows that the compressed token prioritizes discrete anchors, supporting their role as semantic guidance. We further study token-budget scaling, cross-architecture generalization, inference efficiency, robustness to codebook and masking variations, and the information-theoretic properties of the hybrid bottleneck. These results show that combining continuous appearance features with discrete semantic anchors enables effective extreme visual token compression for efficient VLMs.

cs.CV↗

arXiv2026

SafeNexus: Discovering and Steering Modality-Universal Safety Neurons in MLLMs

Although Large Language Models (LLMs) have demonstrated promising safety performance, extending them to Multimodal Large Language Models (MLLMs) exposes a significant gap between expanded multimodal capabilities and existing safety mechanisms. Current defenses remain predominantly confined to specific modal settings, thereby limiting their robustness against broader cross-modal threats. To bridge this gap, we introduce SafeNexus, a cross-modal safety alignment framework that adopts a dedicated neuron-level intervention strategy. First, we formulate a neuron localization paradigm that identifies functionally specialized neurons by characterizing intermediate-layer activation patterns and quantifying their functional salience through importance scoring. Building upon this paradigm, we exploit contrastive data to identify modality-bound safety neurons (BS-Neurons), and validate their role in regulating safety behavior within each modality via targeted suppression. Further cross-modal analysis defines modality-universal safety neurons (US-Neurons) as the shared subset of BS-Neurons identified across individual modalities, serving as the core for defending against harmful cross-modal attacks. We observe that suppressing these neurons substantially degrades safety performance across modalities, while leaving overall utility largely unaffected. Building on these insights, we propose two safety alignment strategies: activation-level safety amplifier and safety neuron calibrator. The proposed strategies enhance model safety through two distinct routes: the former amplifies the activation magnitudes of US-Neurons, while the latter selectively calibrates them via targeted fine-tuning. Extensive experiments demonstrate that our method outperforms prevailing state-of-the-art approaches on safety benchmarks spanning diverse modality combinations, while effectively preserving utility.

cs.CV↗

arXiv2026

Rethinking Text-Based Image Retrieval in Specific Domain

Driven by the rapid advancement of vision-language representation learning, Text-based Image Retrieval (TBIR) has made notable progress. However, existing benchmarks are predominantly constructed on an exclusive single-match assumption between query and images. While effective in general scenarios, this assumption fails to reflect practical system performance in specific domains (e.g., surveillance), where a single query often corresponds to multiple relevant candidate images. To address this limitation, we design a Domain-Specific Multi-Match Text-based Image Retrieval (DSMM-TBIR) data engine. Leveraging this engine, we construct Security Multi-Match TBIR (SecMM-TBIR), a benchmark comprising 50k surveillance images with 200 comprehensive queries. Furthermore, we observe that vanilla contrastive learning in specific domains suffers from severe false negatives, forcing the model to push apart semantically similar pairs and thus degrading retrieval performance. We propose the Semantic-Aware Fine-Tuning (SAFT) framework to address semantic compression in specific domains, which incorporates Semantic-Aware Soft-Label Supervision (SASS) and Intra-modal Structural Distillation (ISD) to establish a promising paradigm for domain-specific TBIR tasks. Experiments across diverse CLIP-like models demonstrate that SAFT yields an average mAP@20 gain of 7.8 points on SecMM-TBIR over standard image-text contrastive (ITC) fine-tuning, while also improving general-domain performance. The entire benchmark will be released to facilitate further research.

cs.CV↗

arXiv2026

ConRub-Med: Reinforcement Learning with Consensus Rubrics for Open-Ended Medical Question Answering

Reinforcement learning with verifiable rewards has been especially effective in mathematics and coding, where answers can be checked automatically. Many open-ended medical questions lack comparably cheap outcome verifiers: responses may be partly correct, incomplete, or contain clinically consequential errors. Rubrics written or validated by physicians offer strong clinical grounding, but involving experts in every instance is costly. Model-generated rubrics make this supervision scalable. We introduce ConRub-Med to preserve useful distinctions as rubric feedback moves from construction to policy optimization. For each prompt, three heterogeneous language models propose atomic criteria independently; a separate model reviews them, retaining only criteria with semantic support from all three generators. Three-State scoring distinguishes correct coverage, missing information, and incorrect claims. Errors receive negative rather than zero credit. When every response in a complete Group Relative Policy Optimization (GRPO) group receives the same final reward, a pairwise judge provides sequence advantages only if both candidate orders agree, without changing the scalar rewards. Groups without ties use vanilla GRPO. In a blinded study matched by question, two medical experts rate panels from the full pipeline as more clinically relevant than panels produced by one generator. Across the evaluated open models, ConRub-Med ranks first on six of nine benchmarks and achieves the highest medical and generalization averages. Using the resulting rubric dataset of 5,166 prompts, it scores $38.98 \pm 1.04$ (mean $\pm$ SD) on HealthBench-Hard, compared with InfiMed-ORBIT's 33.60 with 8,000 samples and 37.30 with 28,000.

cs.CL↗