arXiv subjects

Xiaohan Xu

Publications and source records attributed to Xiaohan Xu.

16 recordsLinked to original sources

arXiv2026

Beyond Symmetric Fusion: Exploiting Task-Dependent Modality Strengths for RGB-Event Small Object Detection

State-of-the-art RGB-Event detectors improve the detection of small, fast-moving objects by combining complementary features from RGB and Event data, yet they typically fuse the two modalities into a unified representation for both localization and classification. Such a task-symmetric design is inconsistent with the intuition that the two modalities should play different roles according to their task-specific strengths. To examine this issue, we conduct a modality-specific evaluation and find that the relative advantage of the two modalities reverses across tasks: Event data are substantially more effective for class-agnostic localization, whereas RGB data provide stronger category evidence within localized target regions. Motivated by this task-dependent asymmetry, we propose an Asymmetric Event-RGB Object Detection Transformer (AERODet). During class-agnostic localization, Scale-wise Uncertainty-aware Reliability Estimation (SURE) calculates the relative reliability of the two modalities from their objectness response heatmaps and accordingly calibrates their contributions when the decoder aggregates multimodal features. Once the candidate boxes are obtained, Task-Decoupled Semantic Refinement (TDSR) decouples classification from localization and uses RGB RoI features for fine-grained classification. Extensive experiments on FRED and NeRDD demonstrate that AERODet achieves state-of-the-art performance. In particular, it surpasses the strongest RGB-Event baseline by 10.7 mAP points on the FRED challenging split.

cs.CV↗

arXiv2026

UniQL: Towards Dialect-Universal Benchmarking for Text-to-SQL

Existing text-to-SQL benchmarks are largely centered on SQLite, making it difficult to evaluate whether models can generalize across heterogeneous SQL dialects. However, real-world database systems differ substantially in syntax, functions, type systems, and execution semantics, so the same natural language intent often requires dialect-specific SQL realizations. We introduce UniQL, a human-verified benchmark for cross-dialect text-to-SQL evaluation. UniQL aligns 1,534 natural language questions with executable SQL annotations across 16 SQL dialects, yielding 24,544 dialect-specific queries. All dialects share the same intents, aligned schemas and database contents, enabling controlled evaluation of dialect generalization. UniQL is constructed through a hybrid pipeline combining database migration, SQL translation, execution-guided verification, iterative rule summarization, and human validation. Experiments on both open-source and closed-source LLMs show that current models remain far from dialect-universal, with substantial performance variation across database systems and limited transfer from SQLite success to other dialects. These findings highlight the need for aligned cross-dialect benchmarks and more dialect-aware text-to-SQL methods. Code and data are available at https://github.com/JerryGao818/UniQL

cs.AI↗

arXiv2026

Enhance Comprehension of Over-the-Counter Drug Instructions for the General Public and Medical Professionals through Visualization Design

Drug instructions are crucial for guiding the rational use of medication. We conduct a visualization design study to enhance the comprehension of over-the-counter (OTC) drug instructions, targeting both the general public and medical professionals. We devise two tailored drug instruction designs for different audience groups through an iterative design process. A controlled user study reveals that our design outperforms traditional text-based instructions in terms of response time and usability, and the availability of two versions is also found to be beneficial. This study also motivates a taxonomy based on a systematic classification of OTC drug instructions sampled from an official drug database, which received positive expert feedback. Finally, this study summarizes a workflow for a visualization design strategy based on our design exploration and user study feedback, which can be generalized to other OTC drug instructions.

cs.HC↗

arXiv2026

The Evolution of Eco-routing under Population Growth: Evidence from Six U.S. Cities

Rapid urban population growth drives car travel demand, increasing transport carbon emissions and posing a critical challenge to sustainable development. Although existing studies have demonstrated that eco-routing can reduce individual emissions, research gaps remain. On the one hand, such personal reductions have a negligible impact on overall emissions, and cannot be simply aggregated to capture the complex effects of large-scale eco-routing. On the other hand, under population growth, the long-term effectiveness of eco-routing, as well as the evolution of its efficiency and traveler route choice, remain underexplored. To address these limitations, this study proposes Time-Only and Time-Carbon user equilibrium (UE) models, integrates them with a demand forecasting method for simulating future network traffic, and designs multi-dimensional metrics to characterize urban dynamics. Using real-world road networks, commuting origin-destination (OD) demand, and population projections under various shared socioeconomic pathways (SSPs) for six representative U.S. cities as a case study, we conduct a comprehensive analysis of urban dynamics across different routing strategies and population sizes. The results reveal that while eco-routing mitigates total emissions, emissions in most cities scale superlinearly with population, a scaling order that remains invariant regardless of routing and construction strategies. Moreover, under population growth, travelers using eco-routing tend to increasingly select shorter routes, giving rise to carbon bottlenecks. A strategy of targeted capacity expansion on these critical bottlenecks (0.46% of links) significantly reduces both emissions (3%) and travel time (28%) without compromising eco-routing efficiency. This study provides a foundation for formulating low-carbon urban transport planning and emission reduction policies.

physics.soc-ph↗

arXiv2026

Urban Congestion Patterns under High Electric Vehicle Penetration: A Case Study of 10 U.S. Cities

With the global energy transition and the rapid penetration of electric vehicles (EVs), the widening travel cost gap between EVs and gasoline vehicles (GVs) increasingly affects commuters' route choices and may reshape urban congestion patterns. Existing research remains in its preliminary exploratory phase. On the one hand, multi-class models do not account for fixed user class scenarios, which may not align with actual commuters; on the other hand, there is a lack of systematic quantitative analysis based on real-world complex road networks across multiple cities. As a result, the congestion effects induced by heterogeneous GV-EV cost structures may be mischaracterized or substantially underestimated. To address these limitations, this paper proposes a multi-user equilibrium (MUE) assignment model for mixed GV-EV traffic, constructs a dual algorithm with convergence guarantees, and designs multi-dimensional evaluation metrics for congestion patterns. Using 10 representative U.S. cities as a case study, this research explores the evolution trends of traffic congestion under different EV penetration scenarios based on real city-level road networks and block-level commuter origin-destination (OD) demand. The results show that full EV penetration reduces average system travel time by 2.27%--10.78% across the 10 cities, with New Orleans achieving the largest reduction (10.78%) and San Francisco the smallest (2.27%), but the effectiveness of alleviating congestion exhibits urban heterogeneity. Moreover, for cities with sufficient network redundancy, benefits are primarily concentrated during the low to medium EV penetration stage (0-0.5), though cities with topological constraints (e.g., San Francisco) show more limited improvements throughout all penetration levels. This paper can provide a foundation for formulating differentiated urban planning and congestion management policies.

eess.SY↗

arXiv2026

Guided Verifier: Collaborative Multimodal Reasoning via Dynamic Process Supervision

Reinforcement Learning (RL) has emerged as a pivotal mechanism for enhancing the complex reasoning capabilities of Multimodal Large Language Models (MLLMs). However, prevailing paradigms typically rely on solitary rollout strategies where the model works alone. This lack of intermediate oversight renders the reasoning process susceptible to error propagation, where early logical deviations cascade into irreversible failures, resulting in noisy optimization signals. In this paper, we propose the \textbf{Guided Verifier} framework to address these structural limitations. Moving beyond passive terminal rewards, we introduce a dynamic verifier that actively co-solves tasks alongside the policy. During the rollout phase, this verifier interacts with the policy model in real-time, detecting inconsistencies and providing directional signals to steer the model toward valid trajectories. To facilitate this, we develop a specialized data synthesis pipeline targeting multimodal hallucinations, constructing \textbf{CoRe} dataset of process-level negatives and \textbf{Co}rrect-guide \textbf{Re}asoning trajectories to train the guided verifier. Extensive experiments on MathVista, MathVerse and MMMU indicate that by allocating compute to collaborative inference and dynamic verification, an 8B-parameter model can achieve strong performance.

cs.CL↗

arXiv2025

BIRD-INTERACT: Re-imagining Text-to-SQL Evaluation for Large Language Models via Lens of Dynamic Interactions

Large language models (LLMs) have demonstrated remarkable performance on single-turn text-to-SQL tasks, but real-world database applications predominantly require multi-turn interactions to handle ambiguous queries, execution errors, and evolving user requirements. Existing multi-turn benchmarks fall short by treating conversation histories as static context or limiting evaluation to read-only operations, failing to reflect production-grade database assistant challenges. We introduce BIRD-INTERACT, a benchmark that restores this realism through: (1) a comprehensive interaction environment coupling each database with a hierarchical knowledge base, metadata files, and a function-driven user simulator, enabling models to solicit clarifications, retrieve knowledge, and recover from errors without human supervision; (2) two evaluation settings consisting of a pre-defined conversational protocol (c-Interact) and an open-ended agentic setting (a-Interact) where models autonomously decide when to query the user simulator or explore the environment; (3) a challenging task suite covering the full CRUD spectrum for business-intelligence and operational use cases, guarded by executable test cases. Each task features ambiguous and follow-up sub-tasks requiring dynamic interaction. The suite comprises BIRD-INTERACT-FULL (600 tasks, up to 11,796 interactions) for comprehensive performance assessment, and BIRD-INTERACT-LITE (300 tasks with simplified databases) for detailed behavioral analysis and rapid method development. Our empirical results highlight BIRD-INTERACT's difficulty: GPT-5 completes only 8.67% of tasks in c-Interact and 17.00% in a-Interact. Analysis via memory grafting and Interaction Test-time Scaling validates the importance of effective interaction for complex, dynamic text-to-SQL tasks.

cs.AI↗

arXiv2025

SWE-SQL: Illuminating LLM Pathways to Solve User SQL Issues in Real-World Applications

Resolution of complex SQL issues persists as a significant bottleneck in real-world database applications. Current Large Language Models (LLMs), while adept at text-to-SQL translation, have not been rigorously evaluated on the more challenging task of debugging SQL issues. To address this gap, we introduce BIRD-CRITIC, a new SQL issue debugging benchmark comprising 530 PostgreSQL tasks (BIRD-CRITIC-PG) and 570 multi-dialect tasks (BIRD-CRITIC-Multi), distilled from authentic user issues and replayed within new environments to facilitate rigorous evaluation. Baseline evaluations underscore the task's complexity, with the leading reasoning model O3-Mini achieving only 38.87% success rate on BIRD-CRITIC-PG and 33.33% on BIRD-CRITIC-Multi. Meanwhile, advancing open-source models for database tasks is crucial for empowering local development while safeguarding data privacy. Therefore, we present Six-Gym (Sql-fIX-Gym), a training environment for elevating open-source model capabilities for SQL issue debugging. This environment leverages SQL-Rewind strategy, which automatically generates executable issue-solution datasets by reverse-engineering issues from verified SQLs. However, popular trajectory-based fine-tuning methods do not explore substantial supervisory signals. We further propose f-Plan Boosting, which extracts high-level debugging plans from SQL solutions, enabling teacher LLMs to produce 73.7% more successful trajectories for training. We integrate these components into an open-source agent, Bird-Fixer. Based on Qwen-2.5-Coder-14B, Bird-Fixer achieves 38.11% success rate on BIRD-CRITIC-PG and 29.65% on BIRD-CRITIC-Multi, surpassing leading proprietary models such as Claude-3.7-Sonnet and GPT-4.1, marking a significant step toward democratizing sophisticated SQL-debugging capabilities. The leaderboard and source code are available: https://bird-critic.github.io/

cs.DB↗

arXiv2024

MMIDR: Teaching Large Language Model to Interpret Multimodal Misinformation via Knowledge Distillation

Automatic detection of multimodal misinformation has gained a widespread attention recently. However, the potential of powerful Large Language Models (LLMs) for multimodal misinformation detection remains underexplored. Besides, how to teach LLMs to interpret multimodal misinformation in cost-effective and accessible way is still an open question. To address that, we propose MMIDR, a framework designed to teach LLMs in providing fluent and high-quality textual explanations for their decision-making process of multimodal misinformation. To convert multimodal misinformation into an appropriate instruction-following format, we present a data augmentation perspective and pipeline. This pipeline consists of a visual information processing module and an evidence retrieval module. Subsequently, we prompt the proprietary LLMs with processed contents to extract rationales for interpreting the authenticity of multimodal misinformation. Furthermore, we design an efficient knowledge distillation approach to distill the capability of proprietary LLMs in explaining multimodal misinformation into open-source LLMs. To explore several research questions regarding the performance of LLMs in multimodal misinformation detection tasks, we construct an instruction-following multimodal misinformation dataset and conduct comprehensive experiments. The experimental findings reveal that our MMIDR exhibits sufficient detection performance and possesses the capacity to provide compelling rationales to support its assessments.

cs.CL↗

arXiv2024

A Survey on Knowledge Distillation of Large Language Models

In the era of Large Language Models (LLMs), Knowledge Distillation (KD) emerges as a pivotal methodology for transferring advanced capabilities from leading proprietary LLMs, such as GPT-4, to their open-source counterparts like LLaMA and Mistral. Additionally, as open-source LLMs flourish, KD plays a crucial role in both compressing these models, and facilitating their self-improvement by employing themselves as teachers. This paper presents a comprehensive survey of KD's role within the realm of LLM, highlighting its critical function in imparting advanced knowledge to smaller models and its utility in model compression and self-improvement. Our survey is meticulously structured around three foundational pillars: \textit{algorithm}, \textit{skill}, and \textit{verticalization} -- providing a comprehensive examination of KD mechanisms, the enhancement of specific cognitive abilities, and their practical implications across diverse fields. Crucially, the survey navigates the intricate interplay between data augmentation (DA) and KD, illustrating how DA emerges as a powerful paradigm within the KD framework to bolster LLMs' performance. By leveraging DA to generate context-rich, skill-specific training data, KD transcends traditional boundaries, enabling open-source models to approximate the contextual adeptness, ethical alignment, and deep semantic insights characteristic of their proprietary counterparts. This work aims to provide an insightful guide for researchers and practitioners, offering a detailed overview of current methodologies in KD and proposing future research directions. Importantly, we firmly advocate for compliance with the legal terms that regulate the use of LLMs, ensuring ethical and lawful application of KD of LLMs. An associated Github repository is available at https://github.com/Tebmer/Awesome-Knowledge-Distillation-of-LLMs.

cs.CL↗

arXiv2024

Leveraging Large Language Models for NLG Evaluation: Advances and Challenges

In the rapidly evolving domain of Natural Language Generation (NLG) evaluation, introducing Large Language Models (LLMs) has opened new avenues for assessing generated content quality, e.g., coherence, creativity, and context relevance. This paper aims to provide a thorough overview of leveraging LLMs for NLG evaluation, a burgeoning area that lacks a systematic analysis. We propose a coherent taxonomy for organizing existing LLM-based evaluation metrics, offering a structured framework to understand and compare these methods. Our detailed exploration includes critically assessing various LLM-based methodologies, as well as comparing their strengths and limitations in evaluating NLG outputs. By discussing unresolved challenges, including bias, robustness, domain-specificity, and unified evaluation, this paper seeks to offer insights to researchers and advocate for fairer and more advanced NLG evaluation techniques.

cs.CL↗

arXiv2023

Re-Reading Improves Reasoning in Large Language Models

To enhance the reasoning capabilities of off-the-shelf Large Language Models (LLMs), we introduce a simple, yet general and effective prompting method, Re2, i.e., \textbf{Re}-\textbf{Re}ading the question as input. Unlike most thought-eliciting prompting methods, such as Chain-of-Thought (CoT), which aim to elicit the reasoning process in the output, Re2 shifts the focus to the input by processing questions twice, thereby enhancing the understanding process. Consequently, Re2 demonstrates strong generality and compatibility with most thought-eliciting prompting methods, including CoT. Crucially, Re2 facilitates a "bidirectional" encoding in unidirectional decoder-only LLMs because the first pass could provide global information for the second pass. We begin with a preliminary empirical study as the foundation of Re2, illustrating its potential to enable "bidirectional" attention mechanisms. We then evaluate Re2 on extensive reasoning benchmarks across 14 datasets, spanning 112 experiments, to validate its effectiveness and generality. Our findings indicate that, with the exception of a few scenarios on vanilla ChatGPT, Re2 consistently enhances the reasoning performance of LLMs through a simple re-reading strategy. Further analyses reveal Re2's adaptability, showing how it can be effectively integrated with different LLMs, thought-eliciting prompting, and ensemble strategies. Our code is available at \url{https://github.com/Tebmer/Rereading-LLM-Reasoning/}

cs.CL↗

arXiv2023

Cross-modal Contrastive Learning for Multimodal Fake News Detection

Automatic detection of multimodal fake news has gained a widespread attention recently. Many existing approaches seek to fuse unimodal features to produce multimodal news representations. However, the potential of powerful cross-modal contrastive learning methods for fake news detection has not been well exploited. Besides, how to aggregate features from different modalities to boost the performance of the decision-making process is still an open question. To address that, we propose COOLANT, a cross-modal contrastive learning framework for multimodal fake news detection, aiming to achieve more accurate image-text alignment. To further improve the alignment precision, we leverage an auxiliary task to soften the loss term of negative samples during the contrast process. A cross-modal fusion module is developed to learn the cross-modality correlations. An attention mechanism with an attention guidance module is implemented to help effectively and interpretably aggregate the aligned unimodal representations and the cross-modality correlations. Finally, we evaluate the COOLANT and conduct a comparative study on two widely used datasets, Twitter and Weibo. The experimental results demonstrate that our COOLANT outperforms previous approaches by a large margin and achieves new state-of-the-art results on the two datasets.

cs.LG↗

arXiv2023

PoKE: Prior Knowledge Enhanced Emotional Support Conversation with Latent Variable

Emotional support conversation (ESC) task can utilize various support strategies to help people relieve emotional distress and overcome the problem they face, which has attracted much attention in these years. However, most state-of-the-art works rely heavily on external commonsense knowledge to infer the mental state of the user in every dialogue round. Although effective, they may suffer from significant human effort, knowledge update and domain change in a long run. Therefore, in this article, we focus on exploring the task itself without using any external knowledge. We find all existing works ignore two significant characteristics of ESC. (a) Abundant prior knowledge exists in historical conversations, such as the responses to similar cases and the general order of support strategies, which has a great reference value for current conversation. (b) There is a one-to-many mapping relationship between context and support strategy, i.e.multiple strategies are reasonable for a single context. It lays a better foundation for the diversity of generations. Taking into account these two key factors, we propose Prior Knowledge Enhanced emotional support model with latent variable, PoKE. The proposed model fully taps the potential of prior knowledge in terms of exemplars and strategy sequence and then utilizes a latent variable to model the one-to-many relationship of strategy. Furthermore, we introduce a memory schema to incorporate the encoded knowledge into decoder. Experiment results on benchmark dataset show that our PoKE outperforms existing baselines on both automatic evaluation and human evaluation. Compared with the model using external knowledge, PoKE still can make a slight improvement in some metrics. Further experiments prove that abundant prior knowledge is conducive to high-quality emotional support, and a well-learned latent variable is critical to the diversity of generations.

cs.CL↗

arXiv2022

Subgraph Neighboring Relations Infomax for Inductive Link Prediction on Knowledge Graphs

Inductive link prediction for knowledge graph aims at predicting missing links between unseen entities, those not shown in training stage. Most previous works learn entity-specific embeddings of entities, which cannot handle unseen entities. Recent several methods utilize enclosing subgraph to obtain inductive ability. However, all these works only consider the enclosing part of subgraph without complete neighboring relations, which leads to the issue that partial neighboring relations are neglected, and sparse subgraphs are hard to be handled. To address that, we propose Subgraph Neighboring Relations Infomax, SNRI, which sufficiently exploits complete neighboring relations from two aspects: neighboring relational feature for node feature and neighboring relational path for sparse subgraph. To further model neighboring relations in a global way, we innovatively apply mutual information (MI) maximization for knowledge graph. Experiments show that SNRI outperforms existing state-of-art methods by a large margin on inductive link prediction task, and verify the effectiveness of exploring complete neighboring relations in a global way to characterize node features and reason on sparse subgraphs.

cs.AI↗

arXiv2021

Investigation of physical dose enhancement in core-shell magnetic gold nanoparticles with TOPAS simulation

The application of metal nanoparticles as sensitization materials is a common strategy that is used to study dose enhancement in radiotherapy. Recent in vitro tests have revealed that magnetic gold nanoparticles can be used in cancer therapy under a magnetic field to enhance the synergistic efficiency in radiotherapy and photothermal therapy. However, magnetic gold nanoparticles have rarely been studied as sensitization materials. In this study, we obtained further results of the sensitization properties of magnetic gold nanoparticles using the Monte Carlo method TOPAS and TOPAS-nBio. We analyzed the properties of magnetic gold nanoparticles in monoenergetic photons and brachytherapy, and we investigated whether the magnetic field contributes to the sensitization process. Our results demonstrated that the dose enhancement factor of the magnetic gold nanoparticles was 16.7% lower than that of gold nanoparticles in a single particle irradiated by monoenergetic photons. In the cell model, the difference was less than 8.1% in the cytoplasm. We revealed that the magnetic field has no detrimental effect on radiosensitization. Moreover, the sensitization properties of magnetic gold nanoparticles in a clinical brachytherapy source have been revealed for the first time.

physics.med-ph↗