arXiv subjects

Jiawei Cheng

Publications and source records attributed to Jiawei Cheng.

12 recordsLinked to original sources

arXiv2026

DALI: LLM-Agent Enhanced Dual-Stream Adaptive Leadership Identification for Group Recommendations

Group recommendation systems play a pivotal role in supporting collective decisions across various contexts, from leisure activities to organizational team-building. Existing group recommendation approaches typically use either handcrafted aggregation rules (e.g. mean, least misery, weighted sum) or neural aggregation models (e.g. attention-based deep learning frameworks), yet both fall short in distinguishing leader-dominated from collaborative groups and often misrepresent true group preferences, especially when a single member disproportionately influences group choices. To address these limitations, we propose the Dual-stream Adaptive Leadership Identification (DALI) framework, which uniquely combines the symbolic reasoning capabilities of Large Language Models (LLMs) with neural network-based representation learning. Specifically, DALI introduces two key innovations: a dynamic rule generation module that autonomously formulates and evolves identification rules through iterative performance feedback, and a neuro-symbolic aggregation mechanism that concurrently employs symbolic reasoning to robustly recognize leadership groups and attention-based neural aggregation to accurately model collaborative group dynamics. Experiments conducted on the Mafengwo travel dataset confirm that DALI significantly improves recommendation accuracy compared to existing frameworks, highlighting its capability to dynamically adapt to complex, real-world group decision environments.

cs.IR↗

arXiv2026

How to Utilize Complementary Vision-Text Information for 2D Structure Understanding

LLMs typically linearize 2D tables into 1D sequences to fit their autoregressive architecture, which weakens row-column adjacency and other layout cues. In contrast, purely visual encoders can capture spatial cues, yet often struggle to preserve exact cell text. Our analysis reveals that these two modalities provide highly distinct information to LLMs and exhibit strong complementarity. However, direct concatenation and other fusion methods yield limited gains and frequently introduce cross-modal interference. To address this issue, we propose DiVA-Former, a lightweight architecture designed to effectively integrate vision and text information. DiVA-Former leverages visual tokens as dynamic queries to distill long textual sequences into digest vectors, thereby effectively exploiting complementary vision--text information. Evaluated across 13 table benchmarks, DiVA-Former improves upon the pure-text baseline by 23.9\% and achieves consistent gains over existing baselines using visual inputs, textual inputs, or a combination of both.

cs.CV↗

arXiv2026

SkyLink: A Large Vision-Language Model Driven Re-ranking Framework for Cross-View UAV geolocalization

Cross-view UAV geolocalization is fundamentally a challenging large-scale image retrieval task, aiming to determine the geographic coordinates of Unmanned Aerial Vehicle (UAV) queries by matching them against an extensive geo-tagged satellite image database. Most existing methods learn separate feature representations for each view and determine the final prediction using naive heuristics to assess feature similarity, thereby neglecting to model the crucial cross-view relationships. In this paper, we propose SkyLink, a novel plug-and-play ranking framework that pioneers joint relational modeling of inter-view relationships to enhance cross-view UAV geolocalization. SkyLink leverages a Large Vision-Language Model (LVLM) to model the intricate visual-semantic relationships between UAV and satellite views, facilitating effective cross-view matching. To further refine the learning process, we introduce a relational-aware loss. It leverages soft labels to provide a more nuanced supervision signal, mitigating the harsh penalty on near-positive pairs. This approach enhances both training stability and the model's discriminative capacity. Extensive experiments conducted across multiple base retrieval architectures and benchmark datasets demonstrate that SkyLink significantly boosts the ranking effectiveness of existing models, consistently achieving superior performance in various challenging scenarios.

cs.CV↗

arXiv2026

DisenReason: Behavior Disentanglement and Latent Reasoning for Shared-Account Sequential Recommendation

Shared-account usage is common on streaming and e-commerce platforms, where multiple users share one account. Existing shared-account sequential recommendation (SSR) methods often assume a fixed number of latent users per account, limiting their ability to adapt to diverse sharing patterns and reducing recommendation accuracy. Recent latent reasoning technique applied in sequential recommendation (SR) generate intermediate embeddings from the user embedding (e.g, last item embedding) to uncover users' potential interests, which inspires us to treat the problem of inferring the number of latent users as generating a series of intermediate embeddings, shifting from inferring preferences behind user to inferring the users behind account. However, the last item cannot be directly used for reasoning in SSR, as it can only represent the behavior of the most recent latent user, rather than the collective behavior of the entire account. To address this, we propose DisenReason, a two-stage reasoning method tailored to SSR. DisenReason combines behavior disentanglement stage from frequency-domain perspective to create a collective and unified account behavior representation, which serves as a pivot for latent user reasoning stage to infer the number of users behind the account. Experiments on four benchmark datasets show that DisenReason consistently outperforms all state-of-the-art baselines across four benchmark datasets, achieving relative improvements of up to 12.56\% in MRR@5 and 6.06\% in Recall@20.

cs.IR↗

arXiv2025

RoSA: Enhancing Parameter-Efficient Fine-Tuning via RoPE-aware Selective Adaptation in Large Language Models

Fine-tuning large language models is essential for task-specific adaptation, yet it remains computationally prohibitive. Parameter-Efficient Fine-Tuning (PEFT) methods have emerged as a solution, but current approaches typically ignore the distinct roles of model components and the heterogeneous importance across layers, thereby limiting adaptation efficiency. Motivated by the observation that Rotary Position Embeddings (RoPE) induce critical activations in the low-frequency dimensions of attention states, we propose RoPE-aware Selective Adaptation (RoSA), a novel PEFT framework that allocates trainable parameters in a more targeted and effective manner. RoSA comprises a RoPE-aware Attention Enhancement (RoAE) module, which selectively enhances the low-frequency components of RoPE-influenced attention states, and a Dynamic Layer Selection (DLS) strategy that adaptively identifies and updates the most critical layers based on LayerNorm gradient norms. By combining dimension-wise enhancement with layer-wise adaptation, RoSA achieves more targeted and efficient fine-tuning. Extensive experiments on fifteen commonsense and arithmetic benchmarks demonstrate that RoSA outperforms existing mainstream PEFT methods under comparable trainable parameters. The code is available to ease reproducibility at https://github.com/Applied-Machine-Learning-Lab/RoSA.

cs.CL↗

arXiv2025

POI-Enhancer: An LLM-based Semantic Enhancement Framework for POI Representation Learning

POI representation learning plays a crucial role in handling tasks related to user mobility data. Recent studies have shown that enriching POI representations with multimodal information can significantly enhance their task performance. Previously, the textual information incorporated into POI representations typically involved only POI categories or check-in content, leading to relatively weak textual features in existing methods. In contrast, large language models (LLMs) trained on extensive text data have been found to possess rich textual knowledge. However leveraging such knowledge to enhance POI representation learning presents two key challenges: first, how to extract POI-related knowledge from LLMs effectively, and second, how to integrate the extracted information to enhance POI representations. To address these challenges, we propose POI-Enhancer, a portable framework that leverages LLMs to improve POI representations produced by classic POI learning models. We first design three specialized prompts to extract semantic information from LLMs efficiently. Then, the Dual Feature Alignment module enhances the quality of the extracted information, while the Semantic Feature Fusion module preserves its integrity. The Cross Attention Fusion module then fully adaptively integrates such high-quality information into POI representations and Multi-View Contrastive Learning further injects human-understandable semantic information into these representations. Extensive experiments on three real-world datasets demonstrate the effectiveness of our framework, showing significant improvements across all baseline representations.

cs.AI↗

arXiv2025

Continuum limit of fourth-order Schr\"{o}dinger equations on the lattice

In this paper, we consider the discrete fourth-order Schr\"{o}dinger equation on the lattice $h\mathbb{Z}^2$. Uniform Strichartz estimates are established by analyzing frequency localized oscillatory integrals with the method of stationary phase and applying Littlewood-Paley inequalities. As an application, we obtain the precise rate of $L^2$ convergence from the solutions of discrete semilinear equations to those of the corresponding equations on the Euclidean plane $\mathbb{R}^2$ in the contimuum limit $h \rightarrow 0$.

math.AP↗

arXiv2024

CMATH: Cross-Modality Augmented Transformer with Hierarchical Variational Distillation for Multimodal Emotion Recognition in Conversation

Multimodal emotion recognition in conversation (MER) aims to accurately identify emotions in conversational utterances by integrating multimodal information. Previous methods usually treat multimodal information as equal quality and employ symmetric architectures to conduct multimodal fusion. However, in reality, the quality of different modalities usually varies considerably, and utilizing a symmetric architecture is difficult to accurately recognize conversational emotions when dealing with uneven modal information. Furthermore, fusing multi-modality information in a single granularity may fail to adequately integrate modal information, exacerbating the inaccuracy in emotion recognition. In this paper, we propose a novel Cross-Modality Augmented Transformer with Hierarchical Variational Distillation, called CMATH, which consists of two major components, i.e., Multimodal Interaction Fusion and Hierarchical Variational Distillation. The former is comprised of two submodules, including Modality Reconstruction and Cross-Modality Augmented Transformer (CMA-Transformer), where Modality Reconstruction focuses on obtaining high-quality compressed representation of each modality, and CMA-Transformer adopts an asymmetric fusion strategy which treats one modality as the central modality and takes others as auxiliary modalities. The latter first designs a variational fusion network to fuse the fine-grained representations learned by CMA- Transformer into a coarse-grained representations. Then, it introduces a hierarchical distillation framework to maintain the consistency between modality representations with different granularities. Experiments on the IEMOCAP and MELD datasets demonstrate that our proposed model outperforms previous state-of-the-art baselines. Implementation codes can be available at https://github.com/ cjw-MER/CMATH.

cs.MM↗

arXiv2024

Sharp dispersive estimates for the wave equation on the 5-dimensional lattice graph

Schultz \cite{S98} proved dispersive estimates for the wave equation on lattice graphs $\mathbb{Z}^d$ for $d=2,3,$ which was extended to $d=4$ in \cite{BCH23}. By Newton polyhedra and the algorithm introduced by Karpushkin \cite{K83}, we further extend the result to $d=5:$ the sharp decay rate of the fundamental solution of the wave equation on $\mathbb{Z}^5$ is $|t|^{-\frac{11}{6}}.$ Moreover, we prove Strichartz estimates and give applications to nonlinear equations.

math.AP↗

arXiv2024

The fourth-order Schr\"{o}dinger equation on lattices

In this paper, we study the fourth-order Schr\"{o}dinger equation \begin{equation*} i \partial_t u + {\Delta}^2 u - \gamma \Delta u = \pm |u|^{s-1}u \end{equation*} on the lattice $\mathbb{Z}^d$ with dimensions $d=1,2$ and parameter $\gamma \in \mathbb{R}$. In order to establish sharp dispersive estimates, we consider the fundamental solution as an oscillatory integral and analyze the Newton polyhedron of its phase function. Furthermore, we prove Strichartz estimates which yield the existence of global solutions to nonlinear equations with small data.

math.AP↗

arXiv2023

The Wave Equation on Lattices and Oscillatory Integrals

In this paper, we establish sharp dispersive estimates for the linear wave equation on the lattice $\mathbb{Z}^d$ with dimension $d=4$. Combining the singularity theory with results in uniform estimates of oscillatory integrals, we prove that the optimal time decay rate of the fundamental solution is of order $|t|^{-\frac{3}{2}}\log |t|$, which is the first extension of P. Schultz's results \cite{S98} in $d=2,3$ to the higher dimension. Moreover, we notice that the Newton polyhedron can be used not only to interpret the decay rates for $d=2,3,4$, but also to study the most degenerate case for all odd $d\geq 3$. Furthermore, we prove $l^p\rightarrow l^q$ estimates as well as Strichartz estimates and give applications to nonlinear wave equations.

math.AP↗

arXiv2019

Region crossing change on surfaces

Region crossing change is a local operation on link diagrams. The behavior of region crossing change on $S^2$ is well understood. In this paper, we study the behavior of (modified) region crossing change on higher genus surfaces.

math.GT↗