arXiv subjects

Weihao Shen

Publications and source records attributed to Weihao Shen.

7 recordsLinked to original sources

arXiv2026

Unpaired Modality-Agnostic Generative Recommendation

Generative Recommendation (GR) formulates recommendation as autoregressive generation over discrete semantic identifiers (IDs). Although recent multimodal GR methods improve semantic ID construction with visual and textual information, they typically require item-level paired observations, restricting tokenization to the intersection of modality availability. Moreover, incorporating unpaired observations is nontrivial because small representation shifts may cross quantization boundaries and produce incompatible identifier sequences. To address this challenge, we propose \textbf{Unpair}ed Modality-Agnostic \textbf{G}enerative \textbf{R}ecommendation (UnpairGR), which learns a unified semantic-ID space from paired, image-only, and text-only observations. UnpairGR confines modality-specific processing to lightweight input projections while sharing the subsequent Transformer and residual codebooks across all observation conditions. Paired observations establish a reliability-guided cross-modal consensus, whereas unimodal observations directly refine the same representations and codes. The learned tokenizer is then fixed to provide stationary targets for a single autoregressive recommender, without feature imputation, modality-specific codebooks, or fallback mappings. Extensive experiments on three benchmark datasets demonstrate that UnpairGR consistently improves recommendation performance under both fully observed and incomplete-observation settings.

cs.IR↗

arXiv2026

Requirement--Evidence Alignment for Compositional E-Commerce Queries

Compositional e-commerce queries express multiple requirements that must hold jointly, yet existing rerankers collapse these constraints into aggregate relevance and often promote topical near misses over feasible products. In this paper, we introduce REAlign, a novel requirement-evidence-aligned reranking framework that explicitly connects typed query requirements with visible evidence. REAlign distinguishes satisfied, violated, and unsupported conditions, constructs requirement-targeted contrasts that expose failure modes, and optimizes duplicate-free partial rankings through Requirement-Aware Group-Relative Policy Optimization. Its list utility preserves relevance while incorporating requirement satisfaction, evidence support, material violations, and output validity. Experiments on two fixed-pool e-commerce benchmarks show consistent improvements over strong supervised and policy-optimization baselines under matched training budgets, with fewer violations among top-ranked candidates and larger gains at shallow ranks. Controlled ablations confirm the complementary value of requirement modeling, evidence grounding, and decomposed optimization.

cs.IR↗

arXiv2026

Mosaic: Multimodal Jailbreak against Closed-Source VLMs via Multi-View Ensemble Optimization

Vision-Language Models (VLMs) are powerful but remain vulnerable to multimodal jailbreak attacks. Existing attacks mainly rely on either explicit visual prompt attacks or gradient-based adversarial optimization. While the former is easier to detect, the latter produces subtle perturbations that are less perceptible, but is usually optimized and evaluated under homogeneous open-source surrogate-target settings, leaving its effectiveness on commercial closed-source VLMs under heterogeneous settings unclear. To examine this issue, we study different surrogate-target settings and observe a consistent gap between homogeneous and heterogeneous settings, a phenomenon we term surrogate dependency. Motivated by this finding, we propose Mosaic, a Multi-view ensemble optimization framework for multimodal jailbreak against closed-source VLMs, which alleviates surrogate dependency under heterogeneous surrogate-target settings by reducing over-reliance on any single surrogate model and visual view. Specifically, Mosaic incorporates three core components: a Text-Side Transformation module, which perturbs refusal-sensitive lexical patterns; a Multi-View Image Optimization module, which updates perturbations under diverse cropped views to avoid overfitting to a single visual view; and a Surrogate Ensemble Guidance module, which aggregates optimization signals from multiple surrogate VLMs to reduce surrogate-specific bias. Extensive experiments on safety benchmarks demonstrate that Mosaic achieves state-of-the-art Attack Success Rate and Average Toxicity against commercial closed-source VLMs.

cs.CV↗

arXiv2026

You Only Anonymize What Is Not Intent-Relevant: Suppressing Non-Intent Privacy Evidence

Anonymizing sensitive information in user text is essential for privacy, yet existing methods often apply uniform treatment across attributes, which can conflict with communicative intent and obscure necessary information. This is particularly problematic when personal attributes are integral to expressive or pragmatic goals. The central challenge lies in determining which attributes to protect, and to what extent, while preserving semantic and pragmatic functions. We propose IntentAnony, a utility-preserving anonymization approach that performs intent-conditioned exposure control. IntentAnony models pragmatic intent and constructs privacy inference evidence chains to capture how distributed cues support attribute inference. Conditioned on intent, it assigns each attribute an exposure budget and selectively suppresses non-intent inference pathways while preserving intent-relevant content, semantic structure, affective nuance, and interactional function. We evaluate IntentAnony using privacy inference success rates, text utility metrics, and human evaluation. The results show an approximately 30% improvement in the overall privacy--utility trade-off, with notably stronger usability of anonymized text compared to prior state-of-the-art methods. Our code is available at https://github.com/Nevaeh7/IntentAnony.

cs.CR↗

arXiv2025

Burger: Robust Graph Denoising-augmentation Fusion and Multi-semantic Modeling in Social Recommendation

In the era of rapid development of social media, social recommendation systems as hybrid recommendation systems have been widely applied. Existing methods capture interest similarity between users to filter out interest-irrelevant relations in social networks that inevitably decrease recommendation accuracy, however, limited research has a focus on the mutual influence of semantic information between the social network and the user-item interaction network for further improving social recommendation. To address these issues, we introduce a social \underline{r}ecommendation model with ro\underline{bu}st g\underline{r}aph denoisin\underline{g}-augmentation fusion and multi-s\underline{e}mantic Modeling(Burger). Specifically, we firstly propose to construct a social tensor in order to smooth the training process of the model. Then, a graph convolutional network and a tensor convolutional network are employed to capture user's item preference and social preference, respectively. Considering the different semantic information in the user-item interaction network and the social network, a bi-semantic coordination loss is proposed to model the mutual influence of semantic information. To alleviate the interference of interest-irrelevant relations on multi-semantic modeling, we further use Bayesian posterior probability to mine potential social relations to replace social noise. Finally, the sliding window mechanism is utilized to update the social tensor as the input for the next iteration. Extensive experiments on three real datasets show Burger has a superior performance compared with the state-of-the-art models.

cs.SI↗

arXiv2023

Giant coercivity induced by perpendicular anisotropy in Mn2.42Fe0.58Sn single crystals

We report the discovery of a giant out-of-plane coercivity in the Fe-doped Mn3Sn single crystals. The compound of Mn2.42Fe0.58Sn exhibits a series of magnetic transitions accompanying with large magnetic anisotropy and electric transport properties. Compared with the ab-plane easy axis in Mn3Sn, it switches to the c-axis in Mn2.42Fe0.58Sn, producing a sufficiently large uniaxial anisotropy. At 2 K, a giant out-of-plane coercivity (Hc) up to 3 T was observed, which originates from the large uniaxial magnetocrystalline anisotropy. The modified Sucksmith-Thompson method was used to determine the values of the second-order and the fourth-order magnetocrystalline anisotropy constants K1 and K2, resulting in values of 6.0 * 104 J/m3 and 4.1 * 105 J/m3 at 2 K, respectively. Even though the Curie temperature (TC) of 200 K for Mn2.42Fe0.58Sn is not high enough for direct application, our research presents a valuable case study of a typical uniaxial anisotropy material.

physics.app-ph↗

arXiv2022

Anomalous resistivity upturn in the van der Waals ferromagnet Fe$_5$GeTe$_2$

Fe$_5$GeTe$_2$ (n = 3, 4, 5) have recently attracted increasing attention due to their two-dimensional van der Waals characteristic and high temperature ferromagnetism, which make promises for spintronic devices. The Fe(1) split site is one important structural characteristic of Fe$_5$GeTe$_2$ which makes it very different from other Fe$_5$GeTe$_2$ (n = 3, 4) systems. The local atomic disorder and short-range order can be induced by the split site. In this work, the high-quality van der Waals ferromagnet Fe$_5$GeTe$_2$ were grown to study the low-temperature transport properties. We found a resistivity upturn below 10 K. The temperature and magnetic field dependence of the resistivity are in good agreement with a combination of the theory of disorder-enhanced three-dimensional electron-electron and single-channel Kondo effect. The Kondo effect exists only at low magnetic field B < 3 T, while electron-electron dominates the appearance for the low-temperature resistivity upturn. We believe that the enhanced three-dimensional electron-electron interaction in this system is induced by the local atomic structural disorder due to the split site of Fe(1). Our results indicate that the split site of Fe plays an important role for the exceptional transport properties.

cond-mat.mtrl-sci↗