arXiv subjects

Xiaochen Wang

Publications and source records attributed to Xiaochen Wang.

At least 19 recordsLinked to original sources

arXiv2026

SG-FSM: A Self-Guiding Zero-Shot Prompting Paradigm for Multi-Hop Question Answering Based on Finite State Machine

Large Language Models with chain-of-thought prompting, such as OpenAI-o1, have shown impressive capabilities in natural language inference tasks. However, Multi-hop Question Answering (MHQA) remains challenging for many existing models due to issues like hallucination, error propagation, and limited context length. To address these challenges and enhance LLMs' performance on MHQA, we propose the Self-Guiding prompting Finite State Machine (SG-FSM), designed to strengthen multi-hop reasoning abilities. Unlike traditional chain-of-thought methods, SG-FSM tackles MHQA by iteratively breaking down complex questions into sub-questions, correcting itself to improve accuracy. It processes one sub-question at a time, dynamically deciding the next step based on the current context and results, functioning much like an automaton. Experiments across various benchmarks demonstrate the effectiveness of our approach, outperforming strong baselines on challenging datasets such as Musique. SG-FSM reduces hallucination, enabling recovery of the correct final answer despite intermediate errors. It also improves adherence to specified output formats, simplifying evaluation significantly.

cs.CL↗

arXiv2026

Lowering topological entropy and asymptotic $h$-expansiveness for amenable group actions

Let $G$ be a countably infinite discrete amenable group acting continuously on a compact metric space $X$. We study the problem of lowering topological entropy over subsets of $X$ and its connections with asymptotic $h$-expansiveness. Let $\{F_n\}$ be a tempered Følner sequence such that $e_G\in F_1\subseteq F_2\subseteq\cdots$ and $|F_n|/\log n\to\infty$. If $(X,G)$ has finite topological entropy, then, for every $h\in[0,h_{\mathrm{top}}(X,G)]$,there exists a non-empty compact set $K_h\subseteq X$ whose topological entropy, Bowen topological entropy and packing topological entropy along $\{F_n\}$ are all equal to $h$. We next consider D-hereditary lowerability, which requires every non-empty analytic (Souslin) subset to be lowerable with respect to Bowen topological entropy. We prove that every subshift over a finite alphabet is D-hereditarily lowerable along an increasing Følner sequence. Combining principal quasi-symbolic extensions with a dimensional entropy inequality for factor maps, we further prove that every asymptotically $h$-expansive $G$-system is D-hereditarily lowerable along an increasing Følner sequence. More generally, suppose that $(X,G)$ has finite topological entropy and tail entropy $a$, and let $\{F_n\}$ be an increasing Følner sequence. If an analytic set $K\subseteq X$ has Bowen topological entropy $H>a$ along $\{F_n\}$, then, for every $h\in[0,H-a]$, there exists a non-empty set $K_h\subseteq K$ such that $h\le h_{\mathrm{top}}^B(K_h,{F_n})\le h+a$. Finally, for tempered increasing Følner sequences satisfying the above growth condition, we prove that asymptotic $h$-expansiveness is equivalent to hereditary uniform lowerability.

math.DS↗

arXiv2026

Multi-Granularity Context-Enhanced RAG over Multimodal Knowledge Graphs

Retrieval-augmented generation (RAG) is widely used to mitigate hallucination issues in large language models (LLMs) and multimodal large language models (MLLMs). In particular, knowledge graph (KG)-based RAG leverages structured knowledge to provide (M)LLMs with high-quality external information. Building on these works, recent studies have explored multimodal knowledge graphs (MMKGs) as knowledge bases for GraphRAG. This enables Graph RAG to integrate knowledge across multiple modalities, thereby further enhancing its performance. However, existing MMKG-based RAG methods generally follow a common pipeline in which different modalities are largely processed independently before being fusion. As a result, textual context is only used to a limited extent during visual information extraction and subsequent multimodal knowledge fusion. This brings a semantic gap between images and text which limits the multimodal GraphRAG performance. To address this issue, we propose a novel framework for constructing a Context-Enhanced MMKG (CEMMKG) to better support multimodal GraphRAG. The proposed CEMMKG enriches each image with complementary textual context at both local and global scopes. Local context goes beyond the surrounding text by incorporating sentences that are semantically related to the image, while global context provides a summary of the entire passage. We further introduce a multi-granularity design for the local context, allowing it to capture semantically relevant information at different levels of detail. Extensive experiments on the selected vision-centric dataset validate that CEMMKG is effective in leveraging contextual information to improve MMKG-based RAG performance. Moreover, its effectiveness across different MMKG-based RAG methods demonstrates its broad applicability.

cs.AI↗

arXiv2026

Hereditary Lowerability of Topological Dynamical Systems

Let $(X,T)$ be a topological dynamical system and let $h(T,K)$ denote the topological entropy of a compact set $K\subset X$. We settle a question and a conjecture raised by Huang, Ye, and Zhang (2014) concerning hereditary lowerability. First, we show that every system with finite topological entropy is hereditarily lowerable: for every nonempty compact set $K\subset X$ and every $0\leq h\leq h(T,K)$, there is a compact set $K_h\subset K$ such that $h(T,K_h)=h$. This gives a negative answer to their Question 2'. Second, we prove that if $(X,T)$ admits an ergodic invariant measure with infinite entropy, then $(X,T)$ is not hereditarily lowerable. More precisely, we construct a compact set $K$ with infinite entropy such that every compact subset of $K$ has entropy either zero or infinity. This proves the conjecture stated immediately after Question 2'.

math.DS↗

arXiv2026

KAMR: Grounding Generation via Knowledge-Aligned Multi-hop Retrieval

Graph-based retrieval-augmented generation increasingly relies on multi-hop retrieval, where answering a query requires composing multiple connected knowledge-graph triplets. However, existing retrievers often rank triplets independently via global semantic matching. Moreover, many multi-hop benchmarks provide only final answers, which limits supervision for query--triplet alignment and causes structurally necessary but weakly aligned facts to be missed. To address these issues, we propose a knowledge-aligned multi-hop retriever, KAMR, which distinguishes anchor triplets that are strongly constrained by the query from connected triplets that are weakly aligned yet structurally linked to the anchors. To mitigate the lack of query-triplet alignment supervision, we build a partial alignment dataset by masking triplet elements and prompting an LLM to generate corresponding queries, and optimize two contrastive objectives for pair-level and element-level matching. At inference time, KAMR retrieves anchors globally and then expands locally to collect connected evidence. Across four benchmarks, three LLM backbones, and fourteen baselines, KAMR consistently improves multi-hop retrieval and downstream question answering performance.

cs.IR↗

arXiv2026

Generalizations of the M&M Game

The MandM Game involves two players who begin with I1 and I2 MandM's. During each round, each player tosses a fair coin: if the coin lands heads, that player eats one MandM, and if it lands tails, the player does not eat. If, at the end of a round, one player still has MandM's while the other has none, then the player with MandM's remaining is declared the winner. If both players eat their last MandM in the same round, the game is said to end in a tie. In [BHM+17], the authors studied the probability of a tie in the MandM Game and derived a simple closed-form expression in the special case where both players start with the same number of MandM's. We generalize the MandM Game in several directions, including allowing players to toss multiple coins per round and modifying the probability distributions of the coin flips. We use the technique of generating functions, Monte Carlo methods, and non-linear curve fitting to study the generalized MandM Game.

math.PR↗

arXiv2026

MKG-RAG-Bench: Benchmarking Retrieval in Multimodal Knowledge Graph-Augmented Generation

Retrieval-augmented generation (RAG) over knowledge graphs has emerged as a promising approach for grounding large language models, yet existing benchmarks largely overlook the challenges of retrieval in multimodal knowledge graph RAG (MKG-RAG). In practice, retrieval is a critical bottleneck: multimodal knowledge is heterogeneous, difficult to align across modalities, and often poorly served by retrievers designed for unstructured corpora. To address this gap, we introduce MKG-RAG-Bench, a cross-domain benchmark explicitly designed to evaluate retrieval in MKG-RAG. MKG-RAG-Bench is constructed from two multimodal knowledge graphs spanning general and medical domains, and includes carefully aligned question-answering datasets that support controlled evaluation of both retrieval and downstream generation. The benchmark is built using an LLM-based curation pipeline that filters low-utility knowledge, generates structurally grounded queries with exact supervision, and systematically covers diverse modality configurations. Through extensive experiments across representative retriever families and modality settings, we show that effective multimodal retrieval remains challenging yet crucial for end-to-end MKG-RAG performance, and that retrieval quality strongly determines generation outcomes. By isolating retrieval as a first-class evaluation target, MKG-RAG-Bench provides a principled foundation for diagnosing current limitations and advancing multimodal knowledge graph RAG systems.

cs.AI↗

arXiv2026

EGTR-Review: Efficient Evidence-Grounded Scientific Peer Review Generation via Multi-Agent Teacher Distillation

Scientific peer review generation has attracted increasing attention for reducing reviewing burdens and providing timely feedback. However, existing Large Language Model (LLM)-based methods often produce generic comments with insufficient evidence support and weak source traceability, while complex multi-agent systems incur high inference costs. To address these challenges, we propose EGTR-Review, an Evidence-Grounded and Traceable Review Generation framework via Multi-Agent Teacher Distillation. EGTR-Review first constructs a multi-agent teacher that performs structure-aware paper decomposition, key-element extraction, external scholarly evidence retrieval, evidence-state labeling, verification reasoning, and review synthesis. It then distills both intermediate reasoning trajectories and final review comments into a lightweight student model through task-prefix-driven multi-task learning. An evidence-weighted objective further reduces the influence of weak, missing, or non-verifiable supervision. Experiments on public peer-review datasets show that EGTR-Review (Student) outperforms strong prompt-based, fine-tuned, and structured/agentic baselines across automatic metrics, LLM-as-Judge evaluation, and human evaluation, while maintaining strong factual grounding and source traceability with substantially lower token consumption and inference time. Our code, prompts, configurations, and sample data are available on GitHub.

cs.CL↗

arXiv2026

Dynamical heterogeneity reverses structural suppression of cooperation

Heterogeneity in individual characteristics and behaviour is a fundamental property of complex dynamical systems. While previous studies on evolutionary dynamics of strategies evolution in various systems have predominantly focused on the structural heterogeneity, dynamical heterogeneity in individuals' strategy update has been largely neglected. Here, we introduce a novel dynamical update mechanism based on individuals' decision-making information, comprising personal and social components. This update rule allows each individual to vary in the weight of personal information and the amount of social information, capturing the general scenario of dynamically heterogeneous populations. We find that cooperation, as a collective prosocial outcome, is significantly enhanced when highly connected individuals on interaction network rely more heavily on personal information and access more social information. This effect is notably absent in homogeneous networks, thereby overturning the prevailing consensus that structural heterogeneity inherently suppresses cooperation. This theoretical prediction is further validated by empirical evidence from GitHub collaboration networks. Furthermore, individuals preferentially linking to those who are well-informed and possess greater personal information further promotes collective cooperation. We additionally reveal that cooperators gain a decisive advantage when relying more on personal information compared to defectors, whereas social information affects cooperators and defectors equivalently. Our findings offer profound insights into how dynamical heterogeneity fundamentally shapes the evolution of collective cooperation in complex systems.

physics.soc-ph↗

arXiv2026

The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution

Real-world language agents must handle complex, multi-step workflows across diverse Apps. For instance, an agent may manage emails by coordinating with calendars and file systems, or monitor a production database to detect anomalies and generate reports following an operating manual. However, existing language agent benchmarks often focus on narrow domains or simplified tasks that lack the diversity, realism, and long-horizon complexity required to evaluate agents' real-world performance. To address this gap, we introduce the Tool Decathlon (dubbed as Toolathlon), a benchmark for language agents offering diverse Apps and tools, realistic environment setup, and reliable execution-based evaluation. Toolathlon spans 32 software applications and 604 tools, ranging from everyday platforms such as Google Calendar and Notion to professional ones like WooCommerce, Kubernetes, and BigQuery. Most of the tools are based on a high-quality set of Model Context Protocol (MCP) servers that we may have revised or implemented ourselves. Unlike prior works, which primarily ensure functional realism but offer limited environment state diversity, we provide realistic initial environment states from real software, such as Canvas courses with dozens of students or real financial spreadsheets. This benchmark includes 108 manually sourced or crafted tasks in total, requiring interacting with multiple Apps over around 20 turns on average to complete. Each task is strictly verifiable through dedicated evaluation scripts. Comprehensive evaluation of SOTA models highlights their significant shortcomings: the best-performing model, Claude-4.5-Sonnet, achieves only a 38.6% success rate with 20.2 tool calling turns on average, while the top open-weights model DeepSeek-V3.2-Exp reaches 20.1%. We expect Toolathlon to drive the development of more capable language agents for real-world, long-horizon task execution.

cs.CL↗

arXiv2025

STSR: High-Fidelity Speech Super-Resolution via Spectral-Transient Context Modeling

Speech super-resolution (SR) reconstructs high-fidelity wideband speech from low-resolution inputs-a task that necessitates reconciling global harmonic coherence with local transient sharpness. While diffusion-based generative models yield impressive fidelity, their practical deployment is often stymied by prohibitive computational demands. Conversely, efficient time-domain architectures lack the explicit frequency representations essential for capturing long-range spectral dependencies and ensuring precise harmonic alignment. We introduce STSR, a unified end-to-end framework formulated in the MDCT domain to circumvent these limitations. STSR employs a Spectral-Contextual Attention mechanism that harnesses hierarchical windowing to adaptively aggregate non-local spectral context, enabling consistent harmonic reconstruction up to 48 kHz. Concurrently, a sparse-aware regularization strategy is employed to mitigate the suppression of transient components inherent in compressed spectral representations. STSR consistently outperforms state-of-the-art baselines in both perceptual fidelity and zero-shot generalization, providing a robust, real-time paradigm for high-quality speech restoration.

cs.SD↗

arXiv2025

Strategy evolution on temporal hypergraphs

Individuals interact and cooperate in structured systems. Many studies represent this structure using static networks, where each link represents a permanent connection between two nodes. However, real interactions are generally not time-invariant and are often not pairwise. Recently, progress has been made in modeling higher-order interactions using hypergraphs, where a link may connect more than two individuals. Here, we study cooperation on temporal hypergraphs, capturing the time-varying, higher-order interactions seen in empirical systems. We find that temporal hypergraphs can promote cooperation compared with static networks, and the latter may even underestimate the cooperation-boosting effects of constrained, local interactions. We further show that cooperation can be facilitated by temporal hypergraphs with sparse components and higher-order interactions. Importantly, when the size of group interactions (hyperedges) is comparable to the population size, relatively small hyperedge sizes best facilitate cooperation. Synthetic and empirical hypergraphs alike affirm our findings, illuminating how temporal, higher-order interactions profoundly shape the evolution of cooperation.

physics.soc-ph↗

arXiv2025

CogSR: Semantic-Aware Speech Super-Resolution via Chain-of-Thought Guided Flow Matching

Applying speech super-resolution (SR) to recordings with severely low sampling rates is a critical challenge in digital archiving and investigative audio recovery. In these scenarios, the input lacks essential acoustic cues. Consequently, existing generative models often fail; without sufficient context, they hallucinate phonetic content, guessing words based on probability rather than meaning. To address this, we propose CogSR, a framework designed specifically for high-precision, offline restoration. Our approach shifts the focus from simple signal mapping to cognitive reconstruction. By integrating a Large Audio-Language Model, we employ Chain-of-Thought reasoning to act as a semantic anchor, while explicit acoustic priors ensure the speaker's identity remains consistent. This guides a Rectified Flow backbone to synthesize high-frequency details that are not only realistic but linguistically accurate. Evaluations show that CogSR effectively eliminates ambiguity in severe degradation regimes, making it a robust solution for restoring high-value legacy and surveillance audio.

cs.SD↗

arXiv2025

Linear Differential Vision Transformer: Learning Visual Contrasts via Pairwise Differentials

Vision Transformers (ViTs) have become a universal backbone for both image recognition and image generation. Yet their Multi-Head Self-Attention (MHSA) layer still performs a quadratic query-key interaction for every token pair, spending the bulk of computation on visually weak or redundant correlations. We introduce Visual-Contrast Attention (VCA), a drop-in replacement for MHSA that injects an explicit notion of discrimination while reducing the theoretical complexity from O(N N C) to O(N n C) with n << N. VCA first distils each head's dense query field into a handful of spatially pooled visual-contrast tokens, then splits them into a learnable positive and negative stream whose differential interaction highlights what truly separates one region from another. The module adds fewer than 0.3M parameters to a DeiT-Tiny backbone, requires no extra FLOPs, and is wholly architecture-agnostic. Empirically, VCA lifts DeiT-Tiny top-1 accuracy on ImageNet-1K from 72.2% to 75.6% (+3.4) and improves three strong hierarchical ViTs by up to 3.1%, while in class-conditional ImageNet generation it lowers FID-50K by 2.1 to 5.2 points across both diffusion (DiT) and flow (SiT) models. Extensive ablations confirm that (i) spatial pooling supplies low-variance global cues, (ii) dual positional embeddings are indispensable for contrastive reasoning, and (iii) combining the two in both stages yields the strongest synergy. VCA therefore offers a simple path towards faster and sharper Vision Transformers. The source code is available at https://github.com/LeapLabTHU/LinearDiff.

cs.CV↗

arXiv2025

Demystifying Numerosity in Diffusion Models -- Limitations and Remedies

Numerosity remains a challenge for state-of-the-art text-to-image generation models like FLUX and GPT-4o, which often fail to accurately follow counting instructions in text prompts. In this paper, we aim to study a fundamental yet often overlooked question: Can diffusion models inherently generate the correct number of objects specified by a textual prompt simply by scaling up the dataset and model size? To enable rigorous and reproducible evaluation, we construct a clean synthetic numerosity benchmark comprising two complementary datasets: GrayCount250 for controlled scaling studies, and NaturalCount6 featuring complex naturalistic scenes. Second, we empirically show that the scaling hypothesis does not hold: larger models and datasets alone fail to improve counting accuracy on our benchmark. Our analysis identifies a key reason: diffusion models tend to rely heavily on the noise initialization rather than the explicit numerosity specified in the prompt. We observe that noise priors exhibit biases toward specific object counts. In addition, we propose an effective strategy for controlling numerosity by injecting count-aware layout information into the noise prior. Our method achieves significant gains, improving accuracy on GrayCount250 from 20.0\% to 85.3\% and on NaturalCount6 from 74.8\% to 86.3\%, demonstrating effective generalization across settings.

cs.CV↗

arXiv2025

Beyond Single Frames: Can LMMs Comprehend Temporal and Contextual Narratives in Image Sequences?

Large Multimodal Models (LMMs) have achieved remarkable success across various visual-language tasks. However, existing benchmarks predominantly focus on single-image understanding, leaving the analysis of image sequences largely unexplored. To address this limitation, we introduce StripCipher, a comprehensive benchmark designed to evaluate capabilities of LMMs to comprehend and reason over sequential images. StripCipher comprises a human-annotated dataset and three challenging subtasks: visual narrative comprehension, contextual frame prediction, and temporal narrative reordering. Our evaluation of 16 state-of-the-art LMMs, including GPT-4o and Qwen2.5VL, reveals a significant performance gap compared to human capabilities, particularly in tasks that require reordering shuffled sequential images. For instance, GPT-4o achieves only 23.93% accuracy in the reordering subtask, which is 56.07% lower than human performance. Further quantitative analysis discuss several factors, such as input format of images, affecting the performance of LLMs in sequential understanding, underscoring the fundamental challenges that remain in the development of LMMs.

cs.CL↗

arXiv2025

EventVAD: Training-Free Event-Aware Video Anomaly Detection

Video Anomaly Detection~(VAD) focuses on identifying anomalies within videos. Supervised methods require an amount of in-domain training data and often struggle to generalize to unseen anomalies. In contrast, training-free methods leverage the intrinsic world knowledge of large language models (LLMs) to detect anomalies but face challenges in localizing fine-grained visual transitions and diverse events. Therefore, we propose EventVAD, an event-aware video anomaly detection framework that combines tailored dynamic graph architectures and multimodal LLMs through temporal-event reasoning. Specifically, EventVAD first employs dynamic spatiotemporal graph modeling with time-decay constraints to capture event-aware video features. Then, it performs adaptive noise filtering and uses signal ratio thresholding to detect event boundaries via unsupervised statistical features. The statistical boundary detection module reduces the complexity of processing long videos for MLLMs and improves their temporal reasoning through event consistency. Finally, it utilizes a hierarchical prompting strategy to guide MLLMs in performing reasoning before determining final decisions. We conducted extensive experiments on the UCF-Crime and XD-Violence datasets. The results demonstrate that EventVAD with a 7B MLLM achieves state-of-the-art (SOTA) in training-free settings, outperforming strong baselines that use 7B or larger MLLMs.

cs.CV↗

arXiv2025

Configurable multi-agent framework for scalable and realistic testing of llm-based agents

Large-language-model (LLM) agents exhibit complex, context-sensitive behaviour that quickly renders static benchmarks and ad-hoc manual testing obsolete. We present Neo, a configurable, multi-agent framework that automates realistic, multi-turn evaluation of LLM-based systems. Neo couples a Question Generation Agent and an Evaluation Agent through a shared context-hub, allowing domain prompts, scenario controls and dynamic feedback to be composed modularly. Test inputs are sampled from a probabilistic state model spanning dialogue flow, user intent and emotional tone, enabling diverse, human-like conversations that adapt after every turn. Applied to a production-grade Seller Financial Assistant chatbot, Neo (i) uncovered edge-case failures across five attack categories with a 3.3% break rate close to the 5.8% achieved by expert human red-teamers, and (ii) delivered 10-12X higher throughput, generating 180 coherent test questions in around 45 mins versus 16h of human effort. Beyond security probing, Neo's stochastic policies balanced topic coverage and conversational depth, yielding broader behavioural exploration than manually crafted scripts. Neo therefore lays a foundation for scalable, self-evolving LLM QA: its agent interfaces, state controller and feedback loops are model-agnostic and extensible to richer factual-grounding and policy-compliance checks. We release the framework to facilitate reproducible, high-fidelity testing of emerging agentic systems.

cs.AI↗