Searcharxiv⌕ Search

arXiv subjects

Jinchuan Zhang

Publications and source records attributed to Jinchuan Zhang.

At least 19 recordsLinked to original sources

ANCHOR: Cluster-Grounded 5W1H Extraction via Large-Small Model Collaboration

Deep semantic understanding of events requires recognizing related events and distinguishing why an event occurs from how it is carried out. 5W1H extraction provides a structured foundation for such understanding by organizing event information around six fundamental questions. However, single-event slot filling provides limited support for these distinctions. We propose ANCHOR, a cluster-grounded framework that combines a large language model (LLM) with a small language model equipped with discriminative heads for accurate 5W1H extraction. The LLM proposes candidate events and generates their initial 4W records, which guide the selection of an event cluster and the subsequent anchoring of its main event. A shared document encoder provides evidence for joint 4W refinement, and two separate sets of boundary and role-scoring heads then extract why and how under event-specific constraints, respectively. Document representations are reused across events and stages, and unsupported roles can remain empty. Following the same principles, an agent-and-skill annotation workflow constructs a benchmark of 26,110 documents from ten source corpora covering three languages and seven domains. We select nine of the resulting datasets to evaluate ANCHOR against strong baselines using four open-weight LLMs. ANCHOR achieves competitive performance. The datasets and code are available at https://github.com/revolution6575/ANCHOR-5W1H.

cs.CL↗

MemCredit: From Relevance to Utility for Decision-Time Memory in Long-Horizon Web Agents

Large language model (LLM) agents increasingly operate over long Web interaction histories, making memory management critical for reliable and efficient decision making. Existing methods primarily optimize storage, retrieval, or compression, but the memories most semantically relevant to the current state are not necessarily those that most improve the agent's action. We introduce MemCredit, a decision-centric memory framework that shifts memory selection from semantic relevance to decision utility. MemCredit estimates the state-, reference-context-, and policy-conditional marginal contribution of individual memories through controlled counterfactual interventions, and distills these intervention-derived credits into a lightweight utility estimator. At inference time, semantic relevance is used only for high-recall candidate generation, while predicted utility arbitrates which memories are retained under a fixed token budget, without modifying the underlying agent policy or requiring additional environment interaction. We evaluate MemCredit on ALFWorld and WebShop against representative memory baselines. On ALFWorld, MemCredit achieves 86.57 SR@50 under a 500-token historical-memory budget, compared with 65.67% for BM25 retrieval. Counterfactual analysis further shows that memory utility is sparse, strongly state dependent, and poorly captured by semantic relevance. These results suggest that decision utility provides a more effective principle than relevance alone for long-horizon agent memory management.

cs.AI↗

TSBPO: Tree-Structured Branching for Policy Optimization in Interactive Language Agents

Terminal feedback in multi-turn agent training leaves both rollout allocation and local credit underdetermined. Dynamics-Predictive Sampling (DPS) [9] provides a task-level predictive sampling reference, but interactive agents also need to choose which visited states merit further continuations. We introduce TSBPO with two method components. A lightweight, fine-grained Root/Branch budget allocator distributes factual rollouts across tasks and additional suffixes across replayable states using likelihood-fitted estimates of terminal-outcome contrast. On the resulting physical rollout trees, a tree-aware step-level credit assignment algorithm compares immediate sibling segments and assigns their local outcome contrasts to the turns they contain. Exact task/observation groups provide a complementary step term. The allocator selects what to sample; observed outcomes determine the subsequent credit. With a Qwen2.5-1.5B actor, fixed checkpoints attain 97.79% seen and 96.64% unseen stochastic success on ALFWorld, and an 81.07% per-episode success rate over six stochastic episodes for each of the 500 WebShop test goals. At the WebShop update-100 checkpoint, removing the observation-group step term yields 75.93% under the same evaluation. These results describe the evaluated training configurations on two interactive benchmarks.

cs.AI↗

I2O2I-Nav: A Benchmark with a Modular Agent for Long-Horizon Indoor-to-Outdoor-to-Indoor Navigation

Service and delivery robots often need to complete long-horizon navigation missions spanning indoor and outdoor environments. Existing embodied navigation benchmarks primarily evaluate isolated stages or partial cross-domain journeys, providing limited support for assessing how heterogeneous navigation capabilities compose under continuous execution. We introduce I2O2I-Nav, a physics-enabled benchmark of 20 real-world composite scenes spanning over 115,000 square meters across multiple buildings and floors, with aligned 3DGS rendering and collision geometry. Its three-level evaluation progresses from independent primitives to two-stage compositions and complete four-stage I2O2I missions. Each stage starts from its predecessor's actual endpoint, while stage-conditional metrics track completion and localize failures along the mission. We further propose the MODULAR-Agent that coordinates frozen navigation policies through completion verification, stage handoffs, and bounded recovery. Evaluation across 1,374 episodes highlights the difficulty of continuous composition. The MODULAR-Agent achieves 40.5-76% success on individual primitives, but full I2O2I success remains low. Compared with a fixed policy chain using the same underlying policies, it more than doubles success on both Indoor-to-Outdoor (I2O) and Outdoor-to-Indoor (O2I), while improving full-mission success from 1.15% to 4.02%. Stage-wise analysis shows that the MODULAR-Agent enables more missions to reach the final object-navigation stage, where conditional success remains low, revealing challenges in both long-horizon stage execution and completion decisions.

cs.RO↗

STAIR: Semantic-Temporal Automaton for Interpretable Reasoning in Temporal Question Answering

By leveraging large-scale pretraining, LLMs can interpret diverse temporal expressions and question formulations without task-specific training. However, existing prompt-based neuro-symbolic systems continue to rely on LLMs for both semantic interpretation and exact temporal inference. Consequently, discrete decisions regarding intervals, time anchors, and ordered states remain vulnerable to probabilistic errors and difficult to verify. We present STAIR, a \textbf{S}emantic-\textbf{T}emporal \textbf{A}utomaton for \textbf{I}nterpretable \textbf{R}easoning. STAIR separates semantic interpretation from precise temporal inference: an answer-free LLM adapter maps complex question formulations to normalized temporal intents, while a deterministic temporal automaton with finite control and guarded transitions executes the corresponding policies over canonicalized evidence. Following a rule-first design, STAIR resolves standard questions without invoking an LLM and applies semantic adaptation only when the rule path fails to produce an executable intent. This approach reduces free-form reasoning, making temporal decisions verifiable and interpretable. Specifically, guarded execution supports precise point-time containment and before/after selection, while semantic adaptation handles non-exact intervals and time-anchored queries. Across the TimeQA-Easy, TimeQA-Hard, TempReason-L2, and TempReason-L3 datasets, STAIR consistently outperforms strong baselines in the TQA task using matched model settings, achieving average F1 improvements of 16.57\% and 3.10\% when utilizing the Qwen2.5-7B and GPT-4o-mini models, respectively. Furthermore, ablations and diagnostic analyses demonstrate that STAIR excels at handling both boundary-sensitive and order-sensitive queries, while its guarded execution and semantic adaptation ensure precise point-time reasoning and inexact intervals, respectively.

cs.CL↗

MA-DAR: Manifold-Aligned Dynamic Adaptive Routing for Continual Temporal Knowledge Graph Reasoning

Continual temporal knowledge graph (TKG) reasoning aims to continuously incorporate newly emerging facts while preserving previously acquired knowledge. Replay-based continual learning has achieved promising performance by revisiting historical representations. However, existing methods primarily focus on what to replay, while largely overlooking how replayed representations should be integrated with current ones. Such direct integration often gives rise to two critical forms of representation conflict: \textit{norm domination} and \textit{semantic blurring}, ultimately degrading continual reasoning performance. To address these challenges, we propose MA-DAR (Manifold-Aligned Dynamic Adaptive Routing), a lightweight plug-and-play framework for replay representation fusion. MA-DAR first aligns replayed and current representations onto a shared manifold to alleviate distribution discrepancies. It then employs a dynamic gating mechanism to learn dimension-wise fusion weights, adaptively determining the contribution of replayed and current representations to the fused representation. Furthermore, a polarization regularizer encourages more decisive routing behaviors by discouraging ambiguous gating decisions, resulting in more stable and effective knowledge integration. Extensive experiments on four public continual TKG benchmarks demonstrate that MA-DAR consistently improves the performance of representative TKG encoders while remaining effective under different replay settings. Comprehensive ablation studies and visualization analyses further verify the effectiveness of manifold alignment and dynamic adaptive routing in mitigating representation conflicts and improving continual reasoning.

cs.LG↗

AdvSerial: Physical Adversarial Attacks on Infrastructure-mounted Pedestrian Detectors via Semantic Feature Suppression

AI-based visual perception systems are increasingly deployed in infrastructure surveillance, including roadside monitoring units, highway cameras, and smart-city pedestrian management systems. The security vulnerability of these systems to physical adversarial attacks poses a direct threat to the reliable operation of transportation infrastructure. We propose AdvSerial, a dynamic 2D--3D joint optimization framework for generating continuous high-angle physical adversarial patches against pedestrian detectors in infrastructure-based scenarios. We UV-map a boundary-aware quilted texture onto 3D garments, combine 2D digital attacks with 3D sparse- and continuous-frame rendering, and explicitly suppress person-specific semantic features while enforcing temporal continuity. A Feature Smooth Quilting strategy reduces visible patch boundaries and bounds cross-seam feature discontinuities. A serial-frame loss encourages long uninterrupted sequences of detection failures. In physical world experiments, AdvSerial achieves a 74.8% attack success rate on YOLO-v5 and degrades mean detection confidence from 84.30% to 39.38%. Experiments spanning eight detectors with different architectures demonstrate strong transferability. Notably, it achieves an $89.71%$ attack success rate on YOLO-v2 and resists both patch-detection defenses (NapGuard) and 3D-temporal perception (Sparse4D-v3). The results reveal persistent, temporally consistent failure modes under high-angle surveillance, and motivate the design of motion-aware and 3D-aware defenses for security-critical infrastructure deployments.

cs.CV↗

ISAP-3D: Identity-Slot Aligned Part-Aware 3D Generation

Part-aware 3D generation aims to synthesize structured objects with semantically meaningful components, yet often suffers from structural ambiguity due to identity-layout entanglement. Existing methods either infer part identity and spatial layout implicitly, which can lead to unstable part allocation (e.g., slot swapping or part merging), or rely on strong layout conditions that are difficult to obtain in practice. We attribute this ambiguity to identity-slot permutation freedom: without explicit identity-slot alignment, the correspondence between semantic parts and generation slots is not identifiable during training, allowing multiple slot assignments to fit the same supervision and leading to inconsistent decomposition. Based on this insight, we argue that stable part-aware generation requires identity-aligned one-to-one slot modelling. We therefore propose an identity-slot aligned framework, ISAP-3D, which anchors each part with semantic identity tokens and performs identity-conditioned one-to-one layout prediction, followed by layout-conditioned geometry synthesis. Structured local-global conditioning maintains identity alignment across semantic, spatial, and geometric stages. We also construct a part-level dataset with a unified semantic protocol to enable learnable and consistent identity-slot alignment. Extensive experiments demonstrate improved structural stability, controllability, and robustness over state-of-the-art part-aware generation baselines.

cs.CV↗

Extracting Events Like Code: A Multi-Agent Programming Framework for Zero-Shot Event Extraction

Zero-shot event extraction (ZSEE) remains a significant challenge for large language models (LLMs) due to the need for complex reasoning and domain-specific understanding. Direct prompting often yields incomplete or structurally invalid outputs--such as misclassified triggers, missing arguments, and schema violations. To address these limitations, we present Agent-Event-Coder (AEC), a novel multi-agent framework that treats event extraction like software engineering: as a structured, iterative code-generation process. AEC decomposes ZSEE into specialized subtasks--retrieval, planning, coding, and verification--each handled by a dedicated LLM agent. Event schemas are represented as executable class definitions, enabling deterministic validation and precise feedback via a verification agent. This programming-inspired approach allows for systematic disambiguation and schema enforcement through iterative refinement. By leveraging collaborative agent workflows, AEC enables LLMs to produce precise, complete, and schema-consistent extractions in zero-shot settings. Experiments across five diverse domains and six LLMs demonstrate that AEC consistently outperforms prior zero-shot baselines, showcasing the power of treating event extraction like code generation. The code and data are released on https://github.com/UESTC-GQJ/Agent-Event-Coder.

cs.CL↗

AdvReal: Physical Adversarial Patch Generation Framework for Security Evaluation of Object Detection Systems

Autonomous vehicles are typical complex intelligent systems with artificial intelligence at their core. However, perception methods based on deep learning are extremely vulnerable to adversarial samples, resulting in security accidents. How to generate effective adversarial examples in the physical world and evaluate object detection systems is a huge challenge. In this study, we propose a unified joint adversarial training framework for both 2D and 3D domains, which simultaneously optimizes texture maps in 2D image and 3D mesh spaces to better address intra-class diversity and real-world environmental variations. The framework includes a novel realistic enhanced adversarial module, with time-space and relighting mapping pipeline that adjusts illumination consistency between adversarial patches and target garments under varied viewpoints. Building upon this, we develop a realism enhancement mechanism that incorporates non-rigid deformation modeling and texture remapping to ensure alignment with the human body's non-rigid surfaces in 3D scenes. Extensive experiment results in digital and physical environments demonstrate that the adversarial textures generated by our method can effectively mislead the target detection model. Specifically, our method achieves an average attack success rate (ASR) of 70.13% on YOLOv12 in physical scenarios, significantly outperforming existing methods such as T-SEA (21.65%) and AdvTexture (19.70%). Moreover, the proposed method maintains stable ASR across multiple viewpoints and distances, with an average attack success rate exceeding 90% under both frontal and oblique views at a distance of 4 meters. This confirms the method's strong robustness and transferability under multi-angle attacks, varying lighting conditions, and real-world distances. The demo video and code can be obtained at https://github.com/Huangyh98/AdvReal.git.

cs.CV↗

AgentAlign: Navigating Safety Alignment in the Shift from Informative to Agentic Large Language Models

The acquisition of agentic capabilities has transformed LLMs from "knowledge providers" to "action executors", a trend that while expanding LLMs' capability boundaries, significantly increases their susceptibility to malicious use. Previous work has shown that current LLM-based agents execute numerous malicious tasks even without being attacked, indicating a deficiency in agentic use safety alignment during the post-training phase. To address this gap, we propose AgentAlign, a novel framework that leverages abstract behavior chains as a medium for safety alignment data synthesis. By instantiating these behavior chains in simulated environments with diverse tool instances, our framework enables the generation of highly authentic and executable instructions while capturing complex multi-step dynamics. The framework further ensures model utility by proportionally synthesizing benign instructions through non-malicious interpretations of behavior chains, precisely calibrating the boundary between helpfulness and harmlessness. Evaluation results on AgentHarm demonstrate that fine-tuning three families of open-source models using our method substantially improves their safety (35.8% to 79.5% improvement) while minimally impacting or even positively enhancing their helpfulness, outperforming various prompting methods. The dataset and code have both been open-sourced.

cs.CR↗

Bridging Generative and Discriminative Learning: Few-Shot Relation Extraction via Two-Stage Knowledge-Guided Pre-training

Few-Shot Relation Extraction (FSRE) remains a challenging task due to the scarcity of annotated data and the limited generalization capabilities of existing models. Although large language models (LLMs) have demonstrated potential in FSRE through in-context learning (ICL), their general-purpose training objectives often result in suboptimal performance for task-specific relation extraction. To overcome these challenges, we propose TKRE (Two-Stage Knowledge-Guided Pre-training for Relation Extraction), a novel framework that synergistically integrates LLMs with traditional relation extraction models, bridging generative and discriminative learning paradigms. TKRE introduces two key innovations: (1) leveraging LLMs to generate explanation-driven knowledge and schema-constrained synthetic data, addressing the issue of data scarcity; and (2) a two-stage pre-training strategy combining Masked Span Language Modeling (MSLM) and Span-Level Contrastive Learning (SCL) to enhance relational reasoning and generalization. Together, these components enable TKRE to effectively tackle FSRE tasks. Comprehensive experiments on benchmark datasets demonstrate the efficacy of TKRE, achieving new state-of-the-art performance in FSRE and underscoring its potential for broader application in low-resource scenarios. \footnote{The code and data are released on https://github.com/UESTC-GQJ/TKRE.

cs.CL↗

Historically Relevant Event Structuring for Temporal Knowledge Graph Reasoning

Temporal Knowledge Graph (TKG) reasoning focuses on predicting events through historical information within snapshots distributed on a timeline. Existing studies mainly concentrate on two perspectives of leveraging the history of TKGs, including capturing evolution of each recent snapshot or correlations among global historical facts. Despite the achieved significant accomplishments, these models still fall short of I) investigating the impact of multi-granular interactions across recent snapshots, and II) harnessing the expressive semantics of significant links accorded with queries throughout the entire history, particularly events exerting a profound impact on the future. These inadequacies restrict representation ability to reflect historical dependencies and future trends thoroughly. To overcome these drawbacks, we propose an innovative TKG reasoning approach towards \textbf{His}torically \textbf{R}elevant \textbf{E}vents \textbf{S}tructuring (HisRES). Concretely, HisRES comprises two distinctive modules excelling in structuring historically relevant events within TKGs, including a multi-granularity evolutionary encoder that captures structural and temporal dependencies of the most recent snapshots, and a global relevance encoder that concentrates on crucial correlations among events relevant to queries from the entire history. Furthermore, HisRES incorporates a self-gating mechanism for adaptively merging multi-granularity recent and historically relevant structuring representations. Extensive experiments on four event-based benchmarks demonstrate the state-of-the-art performance of HisRES and indicate the superiority and effectiveness of structuring historical relevance for TKG reasoning.

cs.LG↗

Efficient second-harmonic emission via strong modal overlap in single-resonant lithium niobate nanocavity

High-efficiency second-harmonic generation (SHG) in compact integrated photonic systems is crucial for advancing nonlinear optical technologies. However, achieving exceptional conversion efficiencies while maintaining stable performance remains a significant challenge. Here, we report a high-Q single-resonant photonic crystal nanobeam cavity (PCNBC) on a polymer-loaded lithium niobate on insulator (LNOI) platform, which enables bright second-harmonic (SH) emission. Through synergistic optimization of modal confinement and spatial overlap in a y-cut LN architecture, our device achieves a normalized SHG conversion efficiency of 163%/W, outperforming previous LN-based photonic crystal cavities LN-based photonic crystal cavities by over three orders of magnitude. The visible SH emission at 768.77 nm exhibits a single-lobe radiation pattern with precise spectral alignment between fundamental (FH) and second-harmonic (SH) modes, a critical feature for integrated photonic circuits. Remarkably, the conversion efficiency remains stable under thermal variations up to 20°C, addressing a key limitation of multi-resonant systems. High-order cavity modes are directly visualized via CCD imaging, confirming strong spatial overlap. This work establishes a record SHG conversion efficiency for LN microcavities and provides a scalable, temperature-insensitive architecture for nonlinear light sources, with immediate applications in quantum optics and chip-scale interconnects.

physics.optics↗

NPC: Neural Predictive Control for Fuel-Efficient Autonomous Trucks

Fuel efficiency is a crucial aspect of long-distance cargo transportation by oil-powered trucks that economize on costs and decrease carbon emissions. Current predictive control methods depend on an accurate model of vehicle dynamics and engine, including weight, drag coefficient, and the Brake-specific Fuel Consumption (BSFC) map of the engine. We propose a pure data-driven method, Neural Predictive Control (NPC), which does not use any physical model for the vehicle. After training with over 20,000 km of historical data, the novel proposed NVFormer implicitly models the relationship between vehicle dynamics, road slope, fuel consumption, and control commands using the attention mechanism. Based on the online sampled primitives from the past of the current freight trip and anchor-based future data synthesis, the NVFormer can infer optimal control command for reasonable fuel consumption. The physical model-free NPC outperforms the base PCC method with 2.41% and 3.45% more significant fuel saving in simulation and open-road highway testing, respectively.

cs.RO↗

Holistic Automated Red Teaming for Large Language Models through Top-Down Test Case Generation and Multi-turn Interaction

Automated red teaming is an effective method for identifying misaligned behaviors in large language models (LLMs). Existing approaches, however, often focus primarily on improving attack success rates while overlooking the need for comprehensive test case coverage. Additionally, most of these methods are limited to single-turn red teaming, failing to capture the multi-turn dynamics of real-world human-machine interactions. To overcome these limitations, we propose HARM (Holistic Automated Red teaMing), which scales up the diversity of test cases using a top-down approach based on an extensible, fine-grained risk taxonomy. Our method also leverages a novel fine-tuning strategy and reinforcement learning techniques to facilitate multi-turn adversarial probing in a human-like manner. Experimental results demonstrate that our framework enables a more systematic understanding of model vulnerabilities and offers more targeted guidance for the alignment process.

cs.CL↗

Learning Granularity Representation for Temporal Knowledge Graph Completion

Temporal Knowledge Graphs (TKGs) incorporate temporal information to reflect the dynamic structural knowledge and evolutionary patterns of real-world facts. Nevertheless, TKGs are still limited in downstream applications due to the problem of incompleteness. Consequently, TKG completion (also known as link prediction) has been widely studied, with recent research focusing on incorporating independent embeddings of time or combining them with entities and relations to form temporal representations. However, most existing methods overlook the impact of history from a multi-granularity aspect. The inherent semantics of human-defined temporal granularities, such as ordinal dates, reveal general patterns to which facts typically adhere. To counter this limitation, this paper proposes \textbf{L}earning \textbf{G}ranularity \textbf{Re}presentation (termed $\mathsf{LGRe}$) for TKG completion. It comprises two main components: Granularity Representation Learning (GRL) and Adaptive Granularity Balancing (AGB). Specifically, GRL employs time-specific multi-layer convolutional neural networks to capture interactions between entities and relations at different granularities. After that, AGB generates adaptive weights for these embeddings according to temporal semantics, resulting in expressive representations of predictions. Moreover, to reflect similar semantics of adjacent timestamps, a temporal loss function is introduced. Extensive experimental results on four event benchmarks demonstrate the effectiveness of $\mathsf{LGRe}$ in learning time-related representations. To ensure reproducibility, our code is available at https://github.com/KcAcoZhang/LGRe.

cs.LG↗

RNG: Reducing Multi-level Noise and Multi-grained Semantic Gap for Joint Multimodal Aspect-Sentiment Analysis

As an important multimodal sentiment analysis task, Joint Multimodal Aspect-Sentiment Analysis (JMASA), aiming to jointly extract aspect terms and their associated sentiment polarities from the given text-image pairs, has gained increasing concerns. Existing works encounter two limitations: (1) multi-level modality noise, i.e., instance- and feature-level noise; and (2) multi-grained semantic gap, i.e., coarse- and fine-grained gap. Both issues may interfere with accurate identification of aspect-sentiment pairs. To address these limitations, we propose a novel framework named RNG for JMASA. Specifically, to simultaneously reduce multi-level modality noise and multi-grained semantic gap, we design three constraints: (1) Global Relevance Constraint (GR-Con) based on text-image similarity for instance-level noise reduction, (2) Information Bottleneck Constraint (IB-Con) based on the Information Bottleneck (IB) principle for feature-level noise reduction, and (3) Semantic Consistency Constraint (SC-Con) based on mutual information maximization in a contrastive learning way for multi-grained semantic gap reduction. Extensive experiments on two datasets validate our new state-of-the-art performance.

cs.CL↗