arXiv subjects

Haonan Zhang

Publications and source records attributed to Haonan Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2025

How does online shopping affect offline price sensitivity?

The rapid rise of e-commerce has transformed consumer behavior, prompting questions about how online adoption influences offline shopping. We examine whether consumers who adopt a retailer's online shopping channels become more price-sensitive in their subsequent offline purchases with that retailer. Using transaction-level data from a large Brazilian pet supplies retailer operating both online and offline, we compare "adopters" -- customers who began shopping online after a period of offline-only purchasing -- with "non-adopters" who remained offline-only. We estimate a discrete choice logit model with individual-level heterogeneity, based on an algorithm that can handle both high-dimensional fixed effects and price endogeneity. We then apply a staggered difference-in-differences approach to the estimated price elasticities and obtain the Average Treatment Effect on the Treated (ATT). We find that offline price sensitivity increases significantly after online adoption in three out of four product categories, particularly in items with low switching costs, such as pet hygiene. These results underscore the importance of recognizing cross-channel effects in consumer behavior and contribute to the literature on pricing and multichannel retailing by identifying online adoption as a key driver of offline price sensitivity.

econ.GN↗

arXiv2025

OpenOmni: Advancing Open-Source Omnimodal Large Language Models with Progressive Multimodal Alignment and Real-Time Self-Aware Emotional Speech Synthesis

Recent advancements in omnimodal learning have significantly improved understanding and generation across images, text, and speech, yet these developments remain predominantly confined to proprietary models. The lack of high-quality omnimodal datasets and the challenges of real-time emotional speech synthesis have notably hindered progress in open-source research. To address these limitations, we introduce \name, a two-stage training framework that integrates omnimodal alignment and speech generation to develop a state-of-the-art omnimodal large language model. In the alignment phase, a pre-trained speech model undergoes further training on text-image tasks, enabling (near) zero-shot generalization from vision to speech, outperforming models trained on tri-modal datasets. In the speech generation phase, a lightweight decoder is trained on speech tasks with direct preference optimization, enabling real-time emotional speech synthesis with high fidelity. Experiments show that \name surpasses state-of-the-art models across omnimodal, vision-language, and speech-language benchmarks. It achieves a 4-point absolute improvement on OmniBench over the leading open-source model VITA, despite using 5x fewer training samples and a smaller model size (7B vs. 7x8B). Additionally, \name achieves real-time speech generation with <1s latency at non-autoregressive mode, reducing inference time by 5x compared to autoregressive methods, and improves emotion classification accuracy by 7.7\%

cs.CL↗

arXiv2025

Imaging a moving point source in R^3 from the time of arrival at sparse observation points

In this paper, we introduce a novel numerical method for reconstructing the trajectory within three-dimensional space, where both the emission moment and spatial location of the point source are unknown. Our approach relies solely on measuring the time of arrival at five or seven properly chosen observation points. By utilizing the distinctive geometric configuration of these five or seven observation points, we establish the uniqueness of the trajectory and emission moment of the point source through rigorous mathematical proofs. Moreover, we analyze the stability of our proposed method. The effectiveness of the method is also verified by numerical experiments.

math.NA↗

arXiv2025

Sticking to the Mean: Detecting Sticky Tokens in Text Embedding Models

Despite the widespread use of Transformer-based text embedding models in NLP tasks, surprising 'sticky tokens' can undermine the reliability of embeddings. These tokens, when repeatedly inserted into sentences, pull sentence similarity toward a certain value, disrupting the normal distribution of embedding distances and degrading downstream performance. In this paper, we systematically investigate such anomalous tokens, formally defining them and introducing an efficient detection method, Sticky Token Detector (STD), based on sentence and token filtering. Applying STD to 40 checkpoints across 14 model families, we discover a total of 868 sticky tokens. Our analysis reveals that these tokens often originate from special or unused entries in the vocabulary, as well as fragmented subwords from multilingual corpora. Notably, their presence does not strictly correlate with model size or vocabulary size. We further evaluate how sticky tokens affect downstream tasks like clustering and retrieval, observing significant performance drops of up to 50%. Through attention-layer analysis, we show that sticky tokens disproportionately dominate the model's internal representations, raising concerns about tokenization robustness. Our findings show the need for better tokenization strategies and model design to mitigate the impact of sticky tokens in future text embedding applications.

cs.CL↗

arXiv2025

High-throughput Super-Resolution Imaging Chip based on Miniaturized Full-frequency Encoded-illumination

A miniaturized full-frequency encoded illumination (mini-FEI) chip is presented for high-throughput super-resolution imaging using the spatial frequency shift (SFS) effect. A tunable full SFS scheme is achieved through propagating and evanescent wave. The multi-illumination modes are precisely and flexibly modulated by an encoded LED array. The light travels to the sample via a set of prisms, producing the super-resolution images with high signal-to-noise ratio (SNR). Mini-FEI super-resolution imaging reaches a resolution of 333 nm (~λ/4NA), close to the theoretical limit, while maintaining a large field of view (FOV) of ~1 mm2. The method is validated on label-free samples including USAF Target, Star Target, and onion root tip cells, all of which could be successfully reconstructed. Through the introduction of integrated LED arrays for evanescent wave excitation, expensive laser systems can be avoided and the system significantly miniaturized. The mini-FEI super-resolution imaging chip is simple and cost effective to fabricate and can be used in conjunction with any inverted brightfield microscope frame and thus has great potential for widespread use in scientific and industrial research environments.

physics.optics↗

arXiv2025

Coding for Quasi-Static Fading Channel with Imperfect CSI at the Transmitter and Quantized Feedback

The classical Schalkwijk-Kailath (SK) scheme for the additive Gaussian noise channel with noiseless feedback is highly efficient since its coding complexity is extremely low and the decoding error doubly exponentially decays as the coding blocklength tends to infinity. However, its application to the fading channel with imperfect CSI at the transmitter (I-CSIT) is challenging since the SK scheme is sensitive to the CSI. In this paper, we investigate how to design SK-type scheme for the quasi-static fading channel with I-CSIT and quantized feedback. By introducing modulo lattice function and an auxiliary signal into the SK-type encoder-decoder of the transceiver, we show that the decoding error caused by the I-CSIT can be perfectly eliminated, resulting in the success of designing SK-type scheme for such a case. The study of this paper provides a way to design efficient coding scheme for fading channels in the presence of imperfect CSI and quantized feedback.

cs.IT↗

arXiv2025

Optimal Feedback Schemes for Dirty Paper Channels With State Estimation at the Receiver

In the literature, it has been shown that feedback does not increase the optimal rate-distortion region of the dirty paper channel with state estimation at the receiver (SE-R). On the other hand, it is well-known that feedback helps to construct low-complexity coding schemes in Gaussian channels, such as the elegant Schalkwijk-Kailath (SK) feedback scheme. This motivates us to explore capacity-achieving SK-type schemes in dirty paper channels with SE-R and feedback. In this paper, we first propose a capacity-achieving feedback scheme for the dirty paper channel with SE-R (DPC-SE-R), which combines the superposition coding and the classical SK-type scheme. Then, we extend this scheme to the dirty paper multiple-access channel with SE-R and feedback, and also show the extended scheme is capacity-achieving. Finally, we discuss how to extend our scheme to a noisy state observation case of the DPC-SE-R. However, the capacity-achieving SK-type scheme for such a case remains unknown.

cs.IT↗

arXiv2025

Seismic resolution enhancement via deep Learning with Knowledge Distillation and Domain Adaptation

High-resolution processing of seismic signals is crucial for subsurface geological characterization and thin-layer reservoir identification. Traditional high-resolution algorithms can partially recover high-frequency information but often lack robustness, computational efficiency, and consideration of inter-trace structural relationships. Many deep learning methods use end-to-end architectures that do not incorporate prior knowledge or address data domain disparities, leading to limited generalization.To overcome these challenges, this paper presents the Domain-Adaptive Knowledge Distillation Network (DAKD-Net), which integrates a knowledge distillation strategy with a domain adaptation mechanism for high-resolution seismic data processing. Trained on datasets from forward modeling, DAKD-Net establishes physical relationships between low and high-resolution data, extracting high-frequency prior knowledge during a guided phase before detail restoration without prior conditions. Domain adaptation enhances the model's generalization to real seismic data, improving both generalization capability and structural expression accuracy.DAKD-Net employs a U-Net backbone to extract spatial structural information from multi-trace seismic profiles. The knowledge distillation mechanism enables prior knowledge transfer, allowing recovery of high-resolution data directly from low-resolution inputs. Domain-adaptive fine-tuning further enhances the network's performance in actual survey areas. Experimental results show that DAKD-Net outperforms traditional methods and classical deep networks in longitudinal resolution and complex structural detail restoration, demonstrating strong robustness and practicality.

physics.geo-ph↗

arXiv2025

LightKG: Efficient Knowledge-Aware Recommendations with Simplified GNN Architecture

Recently, Graph Neural Networks (GNNs) have become the dominant approach for Knowledge Graph-aware Recommender Systems (KGRSs) due to their proven effectiveness. Building upon GNN-based KGRSs, Self-Supervised Learning (SSL) has been incorporated to address the sparity issue, leading to longer training time. However, through extensive experiments, we reveal that: (1)compared to other KGRSs, the existing GNN-based KGRSs fail to keep their superior performance under sparse interactions even with SSL. (2) More complex models tend to perform worse in sparse interaction scenarios and complex mechanisms, like attention mechanism, can be detrimental as they often increase learning difficulty. Inspired by these findings, we propose LightKG, a simple yet powerful GNN-based KGRS to address sparsity issues. LightKG includes a simplified GNN layer that encodes directed relations as scalar pairs rather than dense embeddings and employs a linear aggregation framework, greatly reducing the complexity of GNNs. Additionally, LightKG incorporates an efficient contrastive layer to implement SSL. It directly minimizes the node similarity in original graph, avoiding the time-consuming subgraph generation and comparison required in previous SSL methods. Experiments on four benchmark datasets show that LightKG outperforms 12 competitive KGRSs in both sparse and dense scenarios while significantly reducing training time. Specifically, it surpasses the best baselines by an average of 5.8\% in recommendation accuracy and saves 84.3\% of training time compared to KGRSs with SSL. Our code is available at https://github.com/1371149/LightKG.

cs.IR↗

arXiv2025

Low-Rank Augmented Implicit Neural Representation for Unsupervised High-Dimensional Quantitative MRI Reconstruction

Quantitative magnetic resonance imaging (qMRI) provides tissue-specific parameters vital for clinical diagnosis. Although simultaneous multi-parametric qMRI (MP-qMRI) technologies enhance imaging efficiency, robustly reconstructing qMRI from highly undersampled, high-dimensional measurements remains a significant challenge. This difficulty arises primarily because current reconstruction methods that rely solely on a single prior or physics-informed model to solve the highly ill-posed inverse problem, which often leads to suboptimal results. To overcome this limitation, we propose LoREIN, a novel unsupervised and dual-prior-integrated framework for accelerated 3D MP-qMRI reconstruction. Technically, LoREIN incorporates both low-rank prior and continuity prior via low-rank representation (LRR) and implicit neural representation (INR), respectively, to enhance reconstruction fidelity. The powerful continuous representation of INR enables the estimation of optimal spatial bases within the low-rank subspace, facilitating high-fidelity reconstruction of weighted images. Simultaneously, the predicted multi-contrast weighted images provide essential structural and quantitative guidance, further enhancing the reconstruction accuracy of quantitative parameter maps. Furthermore, our work introduces a zero-shot learning paradigm with broad potential in complex spatiotemporal and high-dimensional image reconstruction tasks, further advancing the field of medical imaging.

eess.IV↗

arXiv2025

OmniCharacter: Towards Immersive Role-Playing Agents with Seamless Speech-Language Personality Interaction

Role-Playing Agents (RPAs), benefiting from large language models, is an emerging interactive AI system that simulates roles or characters with diverse personalities. However, existing methods primarily focus on mimicking dialogues among roles in textual form, neglecting the role's voice traits (e.g., voice style and emotions) as playing a crucial effect in interaction, which tends to be more immersive experiences in realistic scenarios. Towards this goal, we propose OmniCharacter, a first seamless speech-language personality interaction model to achieve immersive RPAs with low latency. Specifically, OmniCharacter enables agents to consistently exhibit role-specific personality traits and vocal traits throughout the interaction, enabling a mixture of speech and language responses. To align the model with speech-language scenarios, we construct a dataset named OmniCharacter-10K, which involves more distinctive characters (20), richly contextualized multi-round dialogue (10K), and dynamic speech response (135K). Experimental results showcase that our method yields better responses in terms of both content and style compared to existing RPAs and mainstream speech-language models, with a response latency as low as 289ms. Code and dataset are available at https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/OmniCharacter.

cs.CL↗

arXiv2025

A dimension-free discrete Remez-type inequality on the polytorus

Consider $f:Ω^n_K \to \mathbf{C}$ a function from the $n$-fold product of multiplicative cyclic groups of order $K$. Any such $f$ may be extended via its Fourier expansion to an analytic polynomial on the polytorus $\mathbf{T}^n$, and the set of such polynomials coincides with the set of all analytic polynomials on $\mathbf{T}^n$ of individual degree at most $K-1$. In this setting it is natural to ask how the supremum norms of $f$ over $\mathbf{T}^n$ and over $Ω_K^n$ compare. We prove the following \emph{discretization of the uniform norm} for low-degree polynomials: if $f$ has degree at most $d$ as an analytic polynomial, then $\|f\|_{\mathbf{T}^n}\leq C(d,K)\|f\|_{Ω_K^n}$ with $C(d,K)$ independent of dimension $n$. As a consequence we also obtain a new proof of the Bohnenblust--Hille inequality for functions on products of cyclic groups. Key to our argument is a special class of Fourier multipliers on $Ω_K^n$ which are $L^\infty\to L^\infty$ bounded independent of dimension when restricted to low-degree polynomials. This class includes projections onto the $k$-homogeneous parts of low-degree polynomials as well as projections of much finer granularity.

math.CA↗

arXiv2025

Spectral Method for 1-D Neutron Transport Equation

In this paper, we present an efficient fully spectral approximation scheme for exploring the one-dimensional steady-state neutron transport equation. Our methodology integrates the spectral-(Petrov-)Galerkin scheme in the spatial dimension with the Legendre-Gauss collocation scheme in the directional dimension. The directional integral in the original problem is discretized with Legendre-Gauss quadrature. We furnish a rigorous proof of the solvability of this scheme and, to our best knowledge, conduct a comprehensive error analysis for the first time. Notably, the order of convergence is optimal in the directional dimension, while in the spatial dimension, it is suboptimal and, importantly, non-improvable. Finally, we verify the computational efficiency and error characteristics of the scheme through several numerical examples.

math.NA↗

arXiv2025

KG4RecEval: Does Knowledge Graph Really Matter for Recommender Systems?

Recommender systems (RSs) are designed to provide personalized recommendations to users. Recently, knowledge graphs (KGs) have been widely introduced in RSs to improve recommendation accuracy. In this study, however, we demonstrate that RSs do not necessarily perform worse even if the KG is downgraded to the user-item interaction graph only (or removed). We propose an evaluation framework KG4RecEval to systematically evaluate how much a KG contributes to the recommendation accuracy of a KG-based RS, using our defined metric KGER (KG utilization efficiency in recommendation). We consider the scenarios where knowledge in a KG gets completely removed, randomly distorted and decreased, and also where recommendations are for cold-start users. Our extensive experiments on four commonly used datasets and a number of state-of-the-art KG-based RSs reveal that: to remove, randomly distort or decrease knowledge does not necessarily decrease recommendation accuracy, even for cold-start users. These findings inspire us to rethink how to better utilize knowledge from existing KGs, whereby we discuss and provide insights into what characteristics of datasets and KG-based RSs may help improve KG utilization efficiency. The code and supplementary material of this paper are available at: https://github.com/HotBento/KG4RecEval.

cs.IR↗

arXiv2024

MMEvol: Empowering Multimodal Large Language Models with Evol-Instruct

The development of Multimodal Large Language Models (MLLMs) has seen significant advancements with increasing demands in various fields (e.g., multimodal agents, embodied intelligence). While model-driven approaches attempt to enhance MLLMs capabilities through diverse architectures, the gains have become increasingly marginal. Conversely, data-driven methods, which scale up image-text instruction data, are more effective but face limited data diversity and complexity challenges. The absence of high-quality data constitutes a significant development barrier for MLLMs. To address the data quality bottleneck, we propose MMEvol, a novel multimodal instruction data evolution framework. This framework iteratively improve data quality through a refined combination of fine-grained perception, cognitive reasoning, and interaction evolution, generating a more complex and diverse image-text instruction dataset that empowers MLLMs with enhanced capabilities. Beginning with an initial set of instructions, SEED-163K, we utilize MMEvol to systematically broaden the diversity of instruction types, extend visual reasoning steps to improve cognitive reasoning abilities, and thoroughly explore fine-grained information within images to enhance visual understanding and robustness. To comprehensively evaluate the effectiveness of our approach, we conduct extensive qualitative analysis and quantitative experiments across 13 vision-language tasks. Compared to baseline models trained with the initial seed data, the results demonstrate that our method achieves an average accuracy improvement of 3.1 percentage points. Furthermore, our approach reaches state-of-the-art (SOTA) performance in nine tasks using significantly less data compared to state-of-the-art models.

cs.CL↗

arXiv2024

Dimension-free discretizations of the uniform norm by small product sets

Let $f$ be an analytic polynomial of degree at most $K-1$. A classical inequality of Bernstein compares the supremum norm of $f$ over the unit circle to its supremum norm over the sampling set of the $K$-th roots of unity. Many extensions of this inequality exist, often understood under the umbrella of Marcinkiewicz-Zygmund-type inequalities for $L^p,1\le p\leq \infty$ norms. We study dimension-free extensions of these discretization inequalities in the high-dimension regime, where existing results construct sampling sets with cardinality growing with the total degree of the polynomial. In this work we show that dimension-free discretizations are possible with sampling sets whose cardinality is independent of $°(f)$ and is instead governed by the maximum individual degree of $f$; i.e., the largest degree of $f$ when viewed as a univariate polynomial in any coordinate. For example, we find that for $n$-variate analytic polynomials $f$ of degree at most $d$ and individual degree at most $K-1$, $\|f\|_{L^\infty(\mathbf{D}^n)}\leq C(X)^d\|f\|_{L^\infty(X^n)}$ for any fixed $X$ in the unit disc $\mathbf{D}$ with $|X|=K$. The dependence on $d$ in the constant is tight for such small sampling sets, which arise naturally for example when studying polynomials of bounded degree coming from functions on products of cyclic groups. As an application we obtain a proof of the cyclic group Bohnenblust-Hille inequality with an explicit constant $O(\log K)^{2d}$.

math.FA↗

arXiv2024

High-fidelity near-diffraction-limited projection through scattering with reference-less transmission matrix

Image projection through scattering media has applications ranging from light delivery through multimode fiber to near-eye displays. Conventional methods utilize the transmission matrix (TM) measured by interfering with a reference beam. However, it is noise-sensitive, often resulting in artifacts that degrade the projection quality. Here we propose to characterize the scattering by computationally retrieving TM from intensity-only measurements and solve the projection problem formulated with the retrieved TM by optimization. We experimentally validate the proposed method by projecting through a multimode fiber. Compared to the conventional methods, it projects improved-quality images with resolution near to the diffraction limit, and simplifies the experimental setup by eliminating the reference. It paves the way for applications of high-quality near-diffraction-limited projection through scattering.

physics.optics↗

arXiv2024

Three lectures on Fourier analysis and learning theory

Fourier analysis on the discrete hypercubes $\{-1,1\}^n$ has found numerous applications in learning theory. A recent breakthrough involves the use of a classical result from Fourier analysis, the Bohnenblust--Hille inequality, in the context of learning low-degree Boolean functions. In these lecture notes, we explore this line of research and discuss recent progress in discrete quantum systems and classical Fourier analysis.

math.FA↗