papers

Publications (166)

cs.CV2019

Flow Guided Short-term Trackers with Cascade Detection for Long-term Tracking

Han Wu, Xueyuan Yang, Yong Yang +1

Object tracking has been studied for decades, but most of the existing works are focused on the short-term tracking. For a long sequence, the object is often fully occluded or out…

physics.geo-ph2020

Weather conditions several hours before the strong earthquake

Tao Chen, Lei Li, Xiao-Xin Zhang +3

Physical phenomena observed before strong earthquake have been reported over centuries. Radon anomalies, electrical signals, water level changes, earthquake lights near the epicent…

cs.CR2020

CoinMagic: A Differential Privacy Framework for Ring Signature Schemes

Wangze Ni, Han Wu, Peng Cheng +4

By allowing users to obscure their transactions via including "mixins" (chaff coins), ring signature schemes have been widely used to protect a sender's identity of a transaction i…

math.NT2025

On Weyl's Subconvex Bound for Cube-Free Hecke characters: Totally Real Case

Olga Balkanova, Dmitry Frolenkov, Han Wu

We prove a Weyl-type subconvex bound for cube-free level Hecke characters over totally real number fields. Our proof relies on an explicit inversion to Motohashi's formula. Schwart…

cs.LG2026

Cold-Start Active Correlation Clustering

Linus Aronsson, Han Wu, Morteza Haghir Chehreghani

We study active correlation clustering where pairwise similarities are not provided upfront and must be queried in a cost-efficient manner through active learning. Specifically, we…

stat.ME2025

Noise-Induced Randomization in Regression Discontinuity Designs

Dean Eckles, Nikolaos Ignatiadis, Stefan Wager +1

Regression discontinuity designs assess causal effects in settings where treatment is determined by whether an observed running variable crosses a pre-specified threshold. Here we…

physics.app-ph2022

Stacked conductive metal organic framework nanorods for high performance vacuum electronic devices

Zhengxin Guan, Jun Li, Han Wu +2

Metal-organic frameworks (MOFs) possessing many unique features have been utilized in several fields in recent years. However, their application in field emission (FE) vacuum elect…

math.NT2022

Bias of Root Numbers for Hilbert Newforms of Cubic Level

Zhilin Luo, Qinghua Pi, Han Wu

We give a general formula of the bias of root numbers for Hilbert modular newforms of cubic level. Explicit calculation is given when the base field is $\mathbb{Q}, \mathbb{Q}(\sqr…

cs.LG2026

Merging Beyond: Streaming LLM Updates via Activation-Guided Rotations

Yuxuan Yao, Haonan Sheng, Qingsong Lv +11

The escalating scale of Large Language Models (LLMs) necessitates efficient adaptation techniques. Model merging has gained prominence for its efficiency and controllability. Howev…

cs.IR2025

RecGPT-V2 Technical Report

Chao Yi, Dian Chen, Gaoyang Guo +32

Large language models (LLMs) have demonstrated remarkable potential in transforming recommender systems from implicit behavioral pattern matching to explicit intent reasoning. Whil…

cs.AI2025

BPP-Search: Enhancing Tree of Thought Reasoning for Mathematical Modeling Problem Solving

Teng Wang, Wing-Yin Yu, Zhenqi He +8

LLMs exhibit advanced reasoning capabilities, offering the potential to transform natural language questions into mathematical models. However, existing open-source datasets in ope…

cond-mat.supr-con2021

Flatband-Induced Itinerant Ferromagnetism in RbCoSe

Jianwei Huang, Zhicai Wang, Hongsheng Pang +8

CoSe (=K,Rb,Cs) is a homologue of the iron-based superconductor, FeSe. From a comprehensive study of RbCoSe via measurements of magnetization, tran…

cs.GR2025

One-shot Face Sketch Synthesis in the Wild via Generative Diffusion Prior and Instruction Tuning

Han Wu, Junyao Li, Kangbo Zhao +3

Face sketch synthesis is a technique aimed at converting face photos into sketches. Existing face sketch synthesis research mainly relies on training with numerous photo-sketch sam…

cs.CL2025

Sens-Merging: Sensitivity-Guided Parameter Balancing for Merging Large Language Models

Shuqi Liu, Han Wu, Bowei He +3

Recent advances in large language models have led to numerous task-specialized fine-tuned variants, creating a need for efficient model merging techniques that preserve specialized…

cs.CV2023

LDRNet: Enabling Real-time Document Localization on Mobile Devices

Han Wu, Holland Qian, Huaming Wu +1

While Identity Document Verification (IDV) technology on mobile devices becomes ubiquitous in modern business operations, the risk of identity theft and fraud is increasing. The id…

math.NT2017

A Note on Spectral Analysis in Automorphic Representation Theory for : II

Han Wu

We generalize Zagiers work on regularized integral to the singular case in the adelic setting. We develop necessary tools of treating various singular cases of regularized triple p…

cs.IR2026

A Survey on Generative Recommendation: Data, Model, and Tasks

Min Hou, Le Wu, Yuxin Liao +6

Recommender systems serve as foundational infrastructure in modern information ecosystems, helping users navigate digital content and discover items aligned with their preferences.…

cs.LG2026

AFABench: A Generic Framework for Benchmarking Active Feature Acquisition

Valter Schütz, Han Wu, Reza Rezvan +2

In many real-world scenarios, acquiring all features of a data instance can be expensive or impractical due to monetary cost, latency, or privacy concerns. Active Feature Acquisiti…

cs.CL2025

Determine-Then-Ensemble: Necessity of Top-k Union for Large Language Model Ensembling

Yuxuan Yao, Han Wu, Mingyang Liu +5

Large language models (LLMs) exhibit varying strengths and weaknesses across different tasks, prompting recent studies to explore the benefits of ensembling models to leverage thei…

math.NT2016

A Note on Spectral Analysis for : I

Han Wu

We establish the Fourier inversion for the smooth vectors in over a number field , using minimal knowledge from automorphic representation t…

cs.CL2025

Activation-Guided Consensus Merging for Large Language Models

Yuxuan Yao, Shuqi Liu, Zehua Liu +6

Recent research has increasingly focused on reconciling the reasoning capabilities of System 2 with the efficiency of System 1. While existing training-based and prompt-based appro…

cs.SI2019

MCNE: An End-to-End Framework for Learning Multiple Conditional Network Representations of Social Network

Hao Wang, Tong Xu, Qi Liu +5

Recently, the Network Representation Learning (NRL) techniques, which represent graph structure via low-dimension vectors to support social-oriented application, have attracted wid…

cs.IT2026

Exponential Error Bounds for Information Bottleneck Source Coding Problems

Han Wu, Hamdi Joudeh

We study the information bottleneck (IB) source coding problem, also known as remote lossy source coding under logarithmic loss. Based on a rate-limited description of noisy observ…

cs.CL2024

Learning From Correctness Without Prompting Makes LLM Efficient Reasoner

Yuxuan Yao, Han Wu, Zhijiang Guo +6

Large language models (LLMs) have demonstrated outstanding performance across various tasks, yet they still exhibit limitations such as hallucination, unfaithful reasoning, and tox…

cs.CV2023

Adversarial Driving: Attacking End-to-End Autonomous Driving

Han Wu, Syed Yunas, Sareh Rowlands +2

As research in deep neural networks advances, deep convolutional networks become promising for autonomous driving tasks. In particular, there is an emerging trend of employing end-…

cs.CV2026

FilmBench: A Film-Grade Benchmark for Cinematic Video Generation

Shengyi Wang, Niantong Li, Guangzheng Hu +27

Progress in video generation keeps narrowing the visual gap between AI-generated and professionally produced footage, yet most benchmarks still draw prompts from web sources or LLM…

math.NT2026

Prime Geodesic Theorem for Arithmetic Compact Surfaces: Principal Congruence Case

Chenhao Tang, Han Wu, Jie Yang +1

We generalize Koyama's bound of the error term in the prime geodesic theorems to the principal congruence subgroups for quaternion algebras. Our method avoids the spectral s…

cs.CV2025

LoViC: Efficient Long Video Generation with Context Compression

Jiaxiu Jiang, Wenbo Li, Jingjing Ren +5

Despite recent advances in diffusion transformers (DiTs) for text-to-video generation, scaling to long-duration content remains challenging due to the quadratic complexity of self-…

cs.CR2022

Federated Learning for Tabular Data: Exploring Potential Risk to Privacy

Han Wu, Zilong Zhao, Lydia Y. Chen +1

Federated Learning (FL) has emerged as a potentially powerful privacy-preserving machine learning methodology, since it avoids exchanging data between participants, but instead exc…

physics.geo-ph2019

Hour-scale persistent negative anomaly of atmospheric electrostatic field near the epicenter before earthquake

Tao Chen, Han Wu, Chi Wang +28

Although earthquake prediction is a big challenge in the world, some simple observational tools can capture many physical signals and demonstrate that an earthquake (EQ) may be for…

physics.optics2022

Radiation build-up and dissipation in random fiber laser

Shengtao Lin, Zinan Wang, Jiaojiao Zhang +3

Random fiber laser (RFL) is a complex physical system that arises from the distributed amplification and the intrinsic stochasticity of the fiber scattering. There has been widespr…

cs.AI2026

From Question Answering to Task Completion: A Survey on Agent System and Harness Design

Jianyuan Guo, Zhiwei Hao, Chengcheng Wang +14

LLM-based agents mark a shift from passive question answering to active task completion: they perceive environments, invoke tools, maintain state, and act over extended horizons. A…

cs.CY2022

Know Your Customer: Balancing Innovation and Regulation for Financial Inclusion

Karen Elliott, Kovila Coopamootoo, Edward Curran +8

Financial inclusion depends on providing adjusted services for citizens with disclosed vulnerabilities. At the same time, the financial industry needs to adhere to a strict regulat…

cs.LG2022

Thompson Sampling with Unrestricted Delays

Han Wu, Stefan Wager

We investigate properties of Thompson Sampling in the stochastic multi-armed bandit problem with delayed feedback. In a setting with i.i.d delays, we establish to our knowledge the…

physics.optics2026

Bridging Photon Statistics and Phase Transitions in Random Fiber Lasers

Yifei Qi, Runhao Li, Jie Li +6

Complex systems exhibit rich equilibrium states, yet the universal principles governing these systems remain unrevealed, motivating the search for novel experimental platforms. Ran…

cs.IT2022

On Joint Communication and Channel Discrimination

Han Wu, Hamdi Joudeh

We consider a basic communication and sensing setup comprising a transmitter, a receiver and a sensor. The transmitter sends an encoded sequence to the receiver through a discrete…

cs.CV2022

FastMIM: Expediting Masked Image Modeling Pre-training for Vision

Jianyuan Guo, Kai Han, Han Wu +3

The combination of transformers and masked image modeling (MIM) pre-training framework has shown great potential in various vision tasks. However, the pre-training computational bu…

cs.CL2025

LoRE-Merging: Exploring Low-Rank Estimation For Large Language Model Merging

Zehua Liu, Han Wu, Yuxuan Yao +4

While most current approaches rely on further training techniques, such as fine-tuning or reinforcement learning, to enhance model capacities, model merging stands out for its abil…

cs.LG2020

Deep Technology Tracing for High-tech Companies

Han Wu, Kun Zhang, Guangyi Lv +5

Technological change and innovation are vitally important, especially for high-tech companies. However, factors influencing their future research and development (R&D) trends are b…

hep-ph2025

Searching for heavy millicharged particles from the atmosphere

Han Wu, Edward Hardy, Ningqiang Song

If millicharged particles (MCPs) exist they can be created in the atmosphere when high energy cosmic rays collide with nuclei and could subsequently be detected at neutrino experim…

cs.CL2023

VCSUM: A Versatile Chinese Meeting Summarization Dataset

Han Wu, Mingjie Zhan, Haochen Tan +3

Compared to news and chat summarization, the development of meeting summarization is hugely decelerated by the limited data. To this end, we introduce a versatile Chinese meeting s…

cs.CL2023

Learning Locality and Isotropy in Dialogue Modeling

Han Wu, Haochen Tan, Mingjie Zhan +4

Existing dialogue modeling methods have achieved promising performance on various dialogue tasks with the aid of Transformer and the large-scale pre-trained language models. Howeve…

cond-mat.str-el2023

Spectral Evidence for Local-Moment Ferromagnetism in van der Waals Metals FeGaTe and FeGeTe

Han Wu, Chaowei Hu, Yaofeng Xie +26

Magnetism in two-dimensional (2D) materials has attracted considerable attention recently for both fundamental understanding of magnetism and their tunability towards device applic…

math.NT2022

Genus one half stacky curves violating the local-global principle

Han Wu, Chang Lv

For any number field, we prove that there exists a stacky curve of genus one half defined over the ring of its integers violating the local-global principle for integral points.

cond-mat.mtrl-sci2025

In-plane polar domains enhanced energy storage

Yu Lei, Xiaoming Shi, Sihan Yan +16

Relaxor ferroelectric thin films are recognized for their ultrahigh power density, rendering them highly promising for energy storage applications in electrical and electronic syst…

math.NT2023

A uniform Weyl bound for L-functions of Hilbert modular forms

Han Wu, Ping Xi

We establish a Weyl-type subconvexity of for spherical Hilbert newforms with level ideal , in which is required to be cube-fr…

physics.optics2024

Mid-infrared computational temporal ghost imaging

Han Wu, Bo Hu, Fei Peng +3

Ghost imaging in the time domain allows for reconstructing fast temporal objects using a slow photodetector. The technique involves correlating random or pre-programmed probing tem…

cs.CV2022

Learning Localization-aware Target Confidence for Siamese Visual Tracking

Jiahao Nie, Han Wu, Zhiwei He +3

Siamese tracking paradigm has achieved great success, providing effective appearance discrimination and size estimation by the classification and regression. While such a paradigm…

cs.LG2025

REG: A Regularization Optimizer for Robust Training Dynamics

Zehua Liu, Han Wu, Xiaojin Fu +4

Optimizers are crucial for the efficient training of Large Language Models (LLMs). While AdamW is the de facto standard, recent structure-aware optimizers like Muon have emerged, w…

cs.LG2025

VT-GAN: Cooperative Tabular Data Synthesis using Vertical Federated Learning

Zilong Zhao, Han Wu, Aad Van Moorsel +1

This paper presents the application of Vertical Federated Learning (VFL) to generate synthetic tabular data using Generative Adversarial Networks (GANs). VFL is a collaborative app…

physics.optics2024

Theoretical demonstration of mode transmission in ZGP-based micrometer waveguide platforms

Siyi Lu, Bo Hu, Xuemei Yang +3

Birefringence phase-matching based \c{hi}(2) ZnGeP2 (ZGP) waveguide platform has been recently reported for excellent mid-infrared laser generation. Here, a detailed theoretical ch…

stat.ME2022

Ensemble Method for Estimating Individualized Treatment Effects

Kevin Wu Han, Han Wu

In many medical and business applications, researchers are interested in estimating individualized treatment effects using data from a randomized experiment. For example in medical…

cond-mat.supr-con2021

Non-Thermal Emergence of an Orbital-Selective Mott Phase in FeTeSe

Jianwei Huang, Rong Yu, Zhijun Xu +15

Electronic correlation is of fundamental importance to high temperature superconductivity. Iron-based superconductors are believed to possess moderate correlation strength, which c…

math.AG2016

Central leaves in loop groups

Eva Viehmann, Han Wu

This paper analyses the finer structure of Newton strata in loop groups. These can be decomposed into so-called central leaves. We define them, and determine their global geometric…

cs.AI2024

A privacy-preserving distributed credible evidence fusion algorithm for collective decision-making

Chaoxiong Ma, Yan Liang, Xinyu Yang +2

The theory of evidence reasoning has been applied to collective decision-making in recent years. However, existing distributed evidence fusion methods lead to participants' prefere…

cs.SI2025

Impacts of Physical-Layer Information on Epidemic Spreading in Cyber-Physical Networked Systems

Xianglai Yuan, Yichao Yao, Han Wu +1

Since Granell et al. proposed a multiplex network for information and epidemic propagation, researchers have explored how information propagation affects epidemic dynamics. However…

cond-mat.str-el2023

Reversible Non-Volatile Electronic Switching in a Near Room Temperature van der Waals Ferromagnet

Han Wu, Lei Chen, Paul Malinowski +37

The ability to reversibly toggle between two distinct states in a non-volatile method is important for information storage applications. Such devices have been realized for phase-c…

math.NT2022

Hybrid subconvexity and the partition function

Nickolas Andersen, Han Wu

We give an upper bound for the error term in the Hardy-Ramanujan-Rademacher formula for the partition function. The main input is a new hybrid subconvexity bound for the central va…

cs.CV2024

SIRST-5K: Exploring Massive Negatives Synthesis with Self-supervised Learning for Robust Infrared Small Target Detection

Yahao Lu, Yupei Lin, Han Wu +3

Single-frame infrared small target (SIRST) detection aims to recognize small targets from clutter backgrounds. Recently, convolutional neural networks have achieved significant adv…

cs.LG2026

MoEMeta: Mixture-of-Experts Meta Learning for Few-Shot Relational Learning

Han Wu, Jie Yin

Few-shot knowledge graph relational learning seeks to perform reasoning over relations given only a limited number of training examples. While existing approaches largely adopt a m…

cs.CL2026

SePO: Self-Evolving Prompt Agent for System Prompt Optimization

Wangcheng Tao, Han Wu, Weng-Fai Wong

System prompt optimization improves agent behavior without modifying the underlying model, yielding human-readable, model-agnostic instructions. Existing methods build a prompt age…

math.NT2022

Non-invariance of weak approximation with Brauer-Manin obstruction for surfaces

Han Wu

In this paper, we study the property of weak approximation with Brauer-Manin obstruction for surfaces with respect to field extensions of number fields. For any nontrivial extensio…

math.AG2026

The Brauer-Manin obstruction on algebraic stacks

Chang Lv, Han Wu

For algebraic stacks over number fields, we define their Brauer-Manin sets, Brauer-Manin pairings, and extend the descent theory of Colliot-Thélène and Sansuc. By extending Sansu…

cs.CR2025

Beyond Statistical Estimation: Differentially Private Individual Computation via Shuffling

Shaowei Wang, Changyu Dong, Xiangfu Song +4

In data-driven applications, preserving user privacy while enabling valuable computations remains a critical challenge. Technologies like differential privacy have been pivotal in…

cs.CV2024

GeminiFusion: Efficient Pixel-wise Multimodal Fusion for Vision Transformer

Ding Jia, Jianyuan Guo, Kai Han +4

Cross-modal transformers have demonstrated superiority in various vision tasks by effectively integrating different modalities. This paper first critiques prior token exchange meth…

cs.AI2023

Adversarial Detection: Attacking Object Detection in Real Time

Han Wu, Syed Yunas, Sareh Rowlands +2

Intelligent robots rely on object detection models to perceive the environment. Following advances in deep learning security it has been revealed that object detection models are v…

cs.CL2021

Domain-Adaptive Pretraining Methods for Dialogue Understanding

Han Wu, Kun Xu, Linfeng Song +3

Language models like BERT and SpanBERT pretrained on open-domain data have obtained impressive gains on various NLP tasks. In this paper, we probe the effectiveness of domain-adapt…

cs.IT2024

Joint Communication and Channel Discrimination

Han Wu, Hamdi Joudeh

We consider a basic joint communication and sensing setup comprising a transmitter, a receiver and a sensor. The transmitter sends a codeword to the receiver through a discrete mem…

eess.IV2024

Geometry-Aware Attenuation Learning for Sparse-View CBCT Reconstruction

Zhentao Liu, Yu Fang, Changjian Li +4

Cone Beam Computed Tomography (CBCT) plays a vital role in clinical imaging. Traditional methods typically require hundreds of 2D X-ray projections to reconstruct a high-quality 3D…

math.NT2022

Explicit subconvexity for

Han Wu

We make the subconvex exponent for cuspidal representation in the work of Michel \& Venkatesh explicit. The result depends on an effective dependence on the `fixed'…

cs.CR2021

Privacy-Preserving Batch-based Task Assignment in Spatial Crowdsourcing with Untrusted Server

Maocheng Li, Jiachuan Wang, Libin Zheng +4

In this paper, we study the privacy-preserving task assignment in spatial crowdsourcing, where the locations of both workers and tasks, prior to their release to the server, are pe…

cs.CV2025

CLIP in Medical Imaging: A Survey

Zihao Zhao, Yuxiao Liu, Han Wu +8

Contrastive Language-Image Pre-training (CLIP), a simple yet effective pre-training paradigm, successfully introduces text supervision to vision models. It has shown promising resu…

cs.RO2024

A Human-in-the-Middle Attack against Object Detection Systems

Han Wu, Sareh Rowlands, Johan Wahlstrom

Object detection systems using deep learning models have become increasingly popular in robotics thanks to the rising power of CPUs and GPUs in embedded systems. However, these mod…

cs.AI2026

RouteMoA: Dynamic Routing without Pre-Inference Boosts Efficient Mixture-of-Agents

Jize Wang, Han Wu, Zhiyuan You +9

Mixture-of-Agents (MoA) improves LLM performance through layered collaboration, but its dense topology raises costs and latency. Existing methods employ LLM judges to filter respon…

cs.CL2023

Reconstruct Before Summarize: An Efficient Two-Step Framework for Condensing and Summarizing Meeting Transcripts

Haochen Tan, Han Wu, Wei Shao +5

Meetings typically involve multiple participants and lengthy conversations, resulting in redundant and trivial content. To overcome these challenges, we propose a two-step framewor…

math.NT2022

Arithmetic of Chatelet surfaces under extensions of base fields

Han Wu

For Châtelet surfaces defined over number fields, we study two arithmetic properties, the Hasse principle and weak approximation, when passing to an extension of the base field. G…

math.NT2021

Châtelet surfaces and non-invariance of the Brauer-Manin obstruction for -folds

Han Wu

In this paper, we construct three kinds of Châtelet surfaces, which have some given arithmetic properties with respect to field extensions of number fields. We then use these cons…

math.AG2026

Comparing local-global obstructions on algebraic stacks

Chang Lv, Han Wu, Xucheng Zhang

For quotient stacks [X/G] with X a smooth quasi-projective geometrically integral k-variety and G a linear k-group, where k is a number field, we extend some relations between loca…

cs.IR2025

ReaSeq: Unleashing World Knowledge via Reasoning for Sequential Modeling

Jiakai Tang, Chuan Wang, Gaoming Yang +31

Industrial recommender systems face two fundamental limitations under the log-driven paradigm: (1) knowledge poverty in ID-based item representations that causes brittle interest m…

math.NT2026

Voronoi-Hankel Transforms

Zhefeng Shen, Han Wu

We prove that the -Fourier transform introduced by Jiang--Luo and the Voronoi--Hankel transform for introduced by the last author are essentially the same one. We give a s…

cond-mat.mtrl-sci2024

Ultralarge polarization in ferroelectric hafnia-based thin films

Han Wu, Kun Lin, Qinghua Zhang +15

Hafnia-based ferroelectrics have become a valuable class of electronic functional materials at the nanoscale, showing great potential for next-generation memory and logic devices.…

math.NT2022

On genus one curves violating the local-global principle

Han Wu

For any number field not containing $\QQ(i),$ we give an explicit construction to prove that there exists an elliptic curve defined over this field such that its Shafarevich-Tate g…

cs.CL2025

Unlocking Efficient Long-to-Short LLM Reasoning with Model Merging

Han Wu, Yuxuan Yao, Shuqi Liu +7

The transition from System 1 to System 2 reasoning in large language models (LLMs) has marked significant advancements in handling complex tasks through deliberate, iterative think…

cs.CV2026

Bridging the Skill Gap in Clinical CBCT Interpretation with CBCTRepD

Qinxin Wu, Fucheng Niu, Hengchuan Zhu +9

Generative AI has advanced rapidly in medical report generation; however, its application to oral and maxillofacial CBCT reporting remains limited, largely because of the scarcity…

cs.CV2025

Adapting Foundation Model for Dental Caries Detection with Dual-View Co-Training

Tao Luo, Han Wu, Tong Yang +2

Accurate dental caries detection from panoramic X-rays plays a pivotal role in preventing lesion progression. However, current detection methods often yield suboptimal accuracy due…

cs.IT2025

Strong Converse Exponent for Remote Lossy Source Coding

Han Wu, Hamdi Joudeh

Past works on remote lossy source coding studied the rate under average distortion and the error exponent of excess distortion probability. In this work, we look into how fast the…

cs.IR2025

An Efficient LLM-based Evolutional Recommendation with Locate-Forget-Update Paradigm

Hao Liu, Le Wu, Min Hou +4

Nowadays, Large Language Models (LLMs) have shown exceptional performance in sequential recommendations, and the adoption of LLM-based recommender systems (LLMRec) is becoming incr…

cs.AI2025

Meta-Semantics Augmented Few-Shot Relational Learning

Han Wu, Jie Yin

Few-shot relational learning on knowledge graph (KGs) aims to perform reasoning over relations with only a few training examples. While current methods have focused primarily on le…

cs.CV2026

RoRA: Role-Oriented Regional Allocation for Visual Token Pruning in MLLMs

Qiyanhui Lu, Han Wu, Rongjian Xu +6

Multimodal large language models (MLLMs) encode images as long visual token sequences, making prefilling and KV-cache storage expensive. Existing training-free pruning methods sele…

math.NT2022

Tate-Shafarevich groups of elliptic curves with nontrivial 2-torsion subgroups

Han Wu

For any number field, we prove that there exists an elliptic curve defined over this field such that its Shafarevich-Tate group has a nontrivial 2-torsion subgroup.

eess.IV2023

Multi-View Vertebra Localization and Identification from CT Images

Han Wu, Jiadong Zhang, Yu Fang +4

Accurately localizing and identifying vertebrae from CT images is crucial for various clinical applications. However, most existing efforts are performed on 3D with cropping patch…

cs.CL2021

CSAGN: Conversational Structure Aware Graph Network for Conversational Semantic Role Labeling

Han Wu, Kun Xu, Linqi Song

Conversational semantic role labeling (CSRL) is believed to be a crucial step towards dialogue understanding. However, it remains a major challenge for existing CSRL parser to hand…

math.NA2026

A Unified Kullback--Leibler Divergence Analysis of Generative Diffusion Models via Entropy Production Rate

Han Wu, Zhiwen Zhang

We introduce a unified framework for the error analysis of generative models based on the entropy production rate of the forward-reverse diffusion process pair. For a pair of conti…

cs.CL2025

Preserving LLM Capabilities through Calibration Data Curation: From Analysis to Optimization

Bowei He, Lihao Yin, Huiling Zhen +5

Post-training compression has been a widely employed approach to scale down large language model (LLM) and facilitate efficient inference. In various proposed compression methods,…

math.NT2021

Non-invariance of the Brauer-Manin obstruction for surfaces

Han Wu

In this paper, we study the properties of weak approximation with Brauer-Manin obstruction and the Hasse principle with Brauer-Manin obstruction for surfaces with respect to field…

cs.CL2022

A Sentence is Worth 128 Pseudo Tokens: A Semantic-Aware Contrastive Learning Framework for Sentence Embeddings

Haochen Tan, Wei Shao, Han Wu +2

Contrastive learning has shown great potential in unsupervised sentence embedding tasks, e.g., SimCSE. However, We find that these existing solutions are heavily affected by superf…

cs.CV2021

Hire-MLP: Vision MLP via Hierarchical Rearrangement

Jianyuan Guo, Yehui Tang, Kai Han +5

Previous vision MLPs such as MLP-Mixer and ResMLP accept linearly flattened image patches as input, making them inflexible for different input sizes and hard to capture spatial inf…

math.NT2018

Subconvex bounds for compact toric integrals

Han Wu

We generalize our method for subconvex bounds for to the setting of the Waldspurger's formula for compact torical integrals. We address the two…

cs.CL2024

Towards Versatile and Efficient Visual Knowledge Integration into Pre-trained Language Models with Cross-Modal Adapters

Xinyun Zhang, Haochen Tan, Han Wu +1

Humans learn language via multi-modal knowledge. However, due to the text-only pre-training scheme, most existing pre-trained language models (PLMs) are hindered from the multi-mod…

cs.CV2021

Distilling Object Detectors via Decoupled Features

Jianyuan Guo, Kai Han, Yunhe Wang +4

Knowledge distillation is a widely used paradigm for inheriting information from a complicated teacher network to a compact student network and maintaining the strong performance.…