Searcharxiv⌕ Search

arXiv subjects

Hong Zhu

Publications and source records attributed to Hong Zhu.

At least 19 recordsLinked to original sources
arXiv2026

Ask, Condition or Abstain: Reinforcement Learning for Missing-Premise Reasoning

Answer-only reinforcement learning (RL) trains reasoning models to solve fully specified problems, but many realistic queries omit a premise needed for a unique answer. In this setting, the useful response is not always refusal: the model should ask for the missing premise, condition its answer on the unknown quantity, or abstain when no informative conditional response is available. We present \emph{Ask-Condition-Abstain Reinforcement Learning} (ACA-RL), a data-augmented RL framework for this setting. Its reasoning-graph-guided pipeline converts well-posed problems into missing-premise training instances with localized gap annotations; ACA-RL then trains on these instances with a structured reward over five observable response behaviors. We also introduce the \emph{Missing-Premise Benchmark} (MPB), a 274-instance human-verified benchmark spanning mathematical, logical, and real-world word problems. Across Qwen3 and Llama models, ACA-RL consistently improves on MPB while preserving competitive performance on well-posed reasoning tasks. Together with the released code, MPB, and training data, this work supports a new mission for NLP evaluation: measuring whether models can recognize when a task is underdetermined and handle uncertainty, not only whether they can answer fully specified questions.

cs.CL↗
arXiv2026

SelectLight: Learning to Select Signal Plans Generated by Distributed Model Predictive Control for Urban Traffic Networks

Coordinated traffic signal control across urban networks must adapt to changing demand while satisfying operational constraints. Multi-objective distributed model predictive control (DMPC) can construct feasible signal plans online, but prescribed rules for selecting among trade-off solutions cannot learn from realized closed-loop outcomes. We propose SelectLight, which implements post-optimization selection by allowing a multi-agent reinforcement learning (MARL) policy to choose directly from plans generated online by DMPC. At each control update, state-pruned multi-objective dynamic programming (SP-MODP) evaluates plans with a Newellian point--spatial queue model and returns a bounded set of mutually nondominated candidate signal plans for total queueing delay, peak queue accumulation, and total number of stops. A topology-aware attention policy trained with independent proximal policy optimization (IPPO) selects one unmodified plan from each variable-size set. This confines learning to candidate selection, preserves the prescribed signal timing constraints, and leaves the selected plan and its predicted objective trade-offs available for inspection. Experiments on two 28-intersection SUMO networks show that SelectLight achieves the best delay-related performance and that its advantage widens with demand. At twice the baseline demand, it reduces queueing delay and waiting time by 5.57% and 6.44%, respectively, relative to the strongest baseline. SelectLight also incurs the lowest transfer loss under every tested demand shift. With a 120 s prediction horizon, the per-intersection 99th-percentile SP-MODP solution time is 5.408 ms, well below the 5 s control interval.

eess.SY↗
arXiv2026

Opti-Agent-Bench: Benchmarking End-to-End Optimization R&D Agents on Real-World Business Problems

LLM-based agents are increasingly deployed to solve optimization problems, yet existing benchmarks evaluate them on pre-structured mathematical formulations that bypass the most critical challenge: translating complex business requirements into correct models and solve efficiently. We introduce Opti-Agent-Bench, an end-to-end benchmark that evaluates Large Language Models (LLMs) across the complete optimization R&D pipeline, from understanding business-language descriptions through mathematical modeling, algorithm selection, and code implementation, to solution report generation. Our design rests on three pillars: (1) businesssemantic authenticity with anti-template traps that defeat pattern matching; (2) modular evaluation with cross-module consistency checking across Problem Understanding, Formal Modeling, Implementation, and Reporting; and (3) the ORAC bi-level validity framework that simultaneously ensures task quality and scoring integrity. Across several industrialscale tasks spanning integer programming, robust optimization, stochastic programming, and non-convex optimization, we expose critical failure modes of current models, including constraint omission, model-code inconsistency, and report-implementation divergence, that remain invisible under conventional single-metric evaluation.

cs.AI↗
arXiv2026

Cross-view Multimodal Vision-Based Assessment Framework for Traditional Chinese Medicine Rehabilitation Training

Vision-based assessment can provide convenient and cost-effective evaluation in Traditional Chinese Medicine (TCM) rehabilitation training, where action quality assessment (AQA) from computer vision offers a promising solution. Existing automatic AQA frameworks for physical therapy typically rely on skeletal data captured from a single viewpoint, which is inefficient for TCM techniques such as acupuncture or Tuina that involve dense hand self-occlusion and complex hand-object interactions. To address these challenges, we propose CME-AQA, a cross-view, multimodal vision-based assessment framework that integrates visual-pose fusion to enhance understanding of environmental context and leverages both first-person and third-person videos during training to improve inference robustness. We collected two dual-view datasets, TCM-AQA61-A (Acupuncture) and TCM-AQA61-T (Tuina), each containing synchronized first-person and third-person recordings of 61 subjects with expert annotations. Experimental results show that our approach achieves superior or comparable mean performance against competitive baselines, achieving over 10% relative improvement in weighted F1 over the best competing method on key rating tasks such as Needle Depth and Quick Needle Insertion, while also reducing mean absolute error in quantitative measures such as insertion time and manipulation frequency. Testing on a CPR dataset further demonstrates comparable performance on several posture-based criteria, suggesting applicability to related structured simulated clinical skill assessments where participant motion is central to evaluation. Overall, CME-AQA enhances assessment accuracy for structured TCM rehabilitation training and facilitates more convenient and effective training-oriented skill evaluation.

cs.CV↗
arXiv2026

FuXi-Linear: Unleashing the Power of Linear Attention in Long-term Time-aware Sequential Recommendation

Modern recommendation systems primarily rely on attention mechanisms with quadratic complexity, which limits their ability to handle long user sequences and slows down inference. While linear attention is a promising alternative, existing research faces three critical challenges: (1) temporal signals are often overlooked or integrated via naive coupling that causes mutual interference between temporal and semantic signals while neglecting behavioral periodicity; (2) insufficient positional information provided by existing linear frameworks; and (3) a primary focus on short sequences and shallow architectures. To address these issues, we propose FuXi-Linear, a linear-complexity model designed for efficient long-sequence recommendation. Our approach introduces two key components: (1) a Temporal Retention Channel that independently computes periodic attention weights using temporal data, preventing crosstalk between temporal and semantic signals; (2) a Linear Positional Channel that integrates positional information through learnable kernels within linear complexity. Moreover, we demonstrate that FuXi-Linear exhibits a robust power-law scaling property at a thousand-length scale, a characteristic largely unexplored in prior linear recommendation studies. Extensive experiments on sequences of several thousand tokens demonstrate that FuXi-Linear outperforms state-of-the-art models in recommendation quality, while achieving up to 10$\times$ speedup in the prefill stage and up to 21$\times$ speedup in the decode stage compared to competitive baselines. Our code has been released in a public repository https://github.com/USTC-StarTeam/fuxi-linear.

cs.IR↗
arXiv2026

Operational Robustness of LLMs on Code Generation

It is now common practice in software development for large language models (LLMs) to be used to generate program code. It is desirable to evaluate the robustness of LLMs for this usage. This paper is concerned in particular with how sensitive LLMs are to variations in descriptions of the coding tasks. However, existing techniques for evaluating this robustness are unsuitable for code generation because the input data space of natural language descriptions is discrete. To address this problem, we propose a robustness evaluation method called scenario domain analysis, which aims to find the expected minimal change in the natural language descriptions of coding tasks that would cause the LLMs to produce incorrect outputs. We have formally proved the theoretical properties of the method and also conducted extensive experiments to evaluate the robustness of four state-of-the-art art LLMs: Gemini-pro, Codex, Llamma2 and Falcon 7B, and have found that we are able to rank these with confidence from best to worst. Moreover, we have also studied how robustness varies in different scenarios, including the variations with the topic of the coding task and with the complexity of its sample solution, and found that robustness is lower for more complex tasks and also lower for more advanced topics, such as multi-threading and data structures.

cs.SE↗
arXiv2026

Robustness of Object Detection of Autonomous Vehicles in Adverse Weather Conditions

As self-driving technology advances toward widespread adoption, determining safe operational thresholds across varying environmental conditions becomes critical for public safety. This paper proposes a method for evaluating the robustness of object detection ML models in autonomous vehicles under adverse weather conditions. It employs data augmentation operators to generate synthetic data that simulates different severance degrees of the adverse operation conditions at progressive intensity levels to find the lowest intensity of the adverse conditions at which the object detection model fails. The robustness of the object detection model is measured by the average first failure coefficients (AFFC) over the input images in the benchmark. The paper reports an experiment with four object detection models: YOLOv5s, YOLOv11s, Faster R-CNN, and Detectron2, utilising seven data augmentation operators that simulate weather conditions fog, rain, and snow, and lighting conditions of dark, bright, flaring, and shadow. The experiment data show that the method is feasible, effective, and efficient to evaluate and compare the robustness of object detection models in various adverse operation conditions. In particular, the Faster R-CNN model achieved the highest robustness with an overall average AFFC of 71.9% over all seven adverse conditions, while YOLO variants showed the AFFC values of 43%. The method is also applied to assess the impact of model training that targets adverse operation conditions using synthetic data on model robustness. It is observed that such training can improve robustness in adverse conditions but may suffer from diminishing returns and forgetting phenomena (i.e., decline in robustness) if overtrained.

cs.CV↗
arXiv2026

RelayGR: Scaling Long-Sequence Generative Recommendation via Cross-Stage Relay-Race Inference

Real-time recommender systems execute multi-stage cascades (retrieval, pre-processing, fine-grained ranking) under strict tail-latency SLOs, leaving only tens of milliseconds for ranking. Generative recommendation (GR) models can improve quality by consuming long user-behavior sequences, but in production their online sequence length is tightly capped by the ranking-stage P99 budget. We observe that the majority of GR tokens encode user behaviors that are independent of the item candidates, suggesting an opportunity to pre-infer a user-behavior prefix once and reuse it during ranking rather than recomputing it on the critical path. Realizing this idea at industrial scale is non-trivial: the prefix cache must survive across multiple pipeline stages before the final ranking instance is determined, the user population implies cache footprints far beyond a single device, and indiscriminate pre-inference would overload shared resources under high QPS. We present RelayGR, a production system that enables in-HBM relay-race inference for GR. RelayGR selectively pre-infers long-term user prefixes, keeps their KV caches resident in HBM over the request lifecycle, and ensures the subsequent ranking can consume them without remote fetches. RelayGR combines three techniques: 1) a sequence-aware trigger that admits only at-risk requests under a bounded cache footprint and pre-inference load, 2) an affinity-aware router that co-locates cache production and consumption by routing both the auxiliary pre-infer signal and the ranking request to the same instance, and 3) a memory-aware expander that uses server-local DRAM to capture short-term cross-request reuse while avoiding redundant reloads. We implement RelayGR on Huawei Ascend NPUs and evaluate it with real queries. Under a fixed P99 SLO, RelayGR supports up to 1.5$\times$ longer sequences and improves SLO-compliant throughput by up to 3.6$\times$.

cs.DC↗
arXiv2025

OFF (Proximal) Newton-type Methods with Inexact Derivatives for Unconstrained Optimization

In this paper, we propose objective-function-free (OFF) variants of the proximal Newton method for nonconvex composite optimization problems and the regularized Newton method for unconstrained optimization problems, respectively, using inexact gradients and Hessians. Theoretical analyses verify that the global and local convergence rates of the proposed algorithms remain consistent with those attained under exact evaluations of the objective function and derivatives. To validate the practical applicability of the theoretical assumptions, a lazy gradient strategy is adopted to provide a verifiable scheme for satisfying the accuracy criteria of approximate gradients. For finite-sum optimization problems, an adaptive sampling strategy is further developed to eliminate the circular dependency between sample size and gradient estimation. The proposed algorithm is proven to achieve locally superlinear and quadratic convergence in expectation. Furthermore, we present an OFF regularized Newton and negative curvature algorithm, which uses inexact derivatives to locate approximate second-order stationary points in unconstrained optimization. The worst-case iteration and (sample) operation complexity of the developed algorithm is consistent with the optimal results in the existing literature.

math.OC↗
arXiv2025

MASTEST: A LLM-Based Multi-Agent System For RESTful API Tests

Testing RESTful API is increasingly important in quality assurance of cloud-native applications. Recent advances in machine learning (ML) techniques have demonstrated that various testing activities can be performed automatically by large language models (LLMs) with reasonable accuracy. This paper develops a multi-agent system called MASTEST that combines LLM-based and programmed agents to form a complete tool chain that covers the whole workflow of API test starting from generating unit and system test scenarios from API specification in the OpenAPI Swagger format, to generating of Pytest test scripts, executing test scripts to interact with web services, to analysing web service response messages to determine test correctness and calculate test coverage. The system also supports the incorporation of human testers in reviewing and correcting LLM generated test artefacts to ensure the quality of testing activities. MASTEST system is evaluated on two LLMs, GPT-4o and DeepSeek V3.1 Reasoner with five public APIs. The performances of LLMs on various testing activities are measured by a wide range of metrics, including unit and system test scenario coverage and API operation coverage for the quality of generated test scenarios, data type correctness, status code coverage and script syntax correctness for the quality of LLM generated test scripts, as well as bug detection ability and usability of LLM generated test scenarios and scripts. Experiment results demonstrated that both DeepSeek and GPT-4o achieved a high overall performance. DeepSeek excels in data type correctness and status code detection, while GPT-4o performs best in API operation coverage. For both models, LLM generated test scripts maintained 100\% syntax correctness and only required minimal manual edits for semantic correctness. These findings indicate the effectiveness and feasibility of MASTEST.

cs.SE↗
arXiv2025

Raman Spectroscopy Insights into the Reorientational Dynamics of Polyanions in Solid Electrolytes

The interaction between cation diffusion and polyanion rotational dynamics has garnered significant attention, yet characterizing the rotational dynamics of polyanions remains challenging. Previous studies have primarily relied on complex techniques such as Nuclear Magnetic Resonance and Quasi-Elastic Neutron Scattering. In this work, we use ab initio molecular dynamics (AIMD) simulations and temperature-dependent Raman spectroscopy to investigate the reorientational dynamics of the NO2- polyanion in Na3ONO2 and NaNO2. Our findings reveal distinct reorientational behaviors and establish a clear correlation between molecular reorientation and Raman spectral features in polyanion-based solid electrolytes. AIMD simulations show that NO2- rotates easily in Na3ONO2, while its rotational dynamics are hindered in NaNO2. Raman spectroscopy confirms these results, with temperature-dependent shifts in the internal vibrational modes of the polyanion. In Na3ONO2, the full width at half maximum (FWHM) of these modes increases with temperature, while NaNO2 exhibits a constant FWHM until a significant jump at the ordered-disordered phase transition. These insights enhance our understanding of polyanion reorientation and offer a framework for characterizing similar dynamics in other solid electrolytes.

cond-mat.mtrl-sci↗
arXiv2025

SimpleVSF: VLM-Scoring Fusion for Trajectory Prediction of End-to-End Autonomous Driving

End-to-end autonomous driving has emerged as a promising paradigm for achieving robust and intelligent driving policies. However, existing end-to-end methods still face significant challenges, such as suboptimal decision-making in complex scenarios. In this paper,we propose SimpleVSF (Simple VLM-Scoring Fusion), a novel framework that enhances end-to-end planning by leveraging the cognitive capabilities of Vision-Language Models (VLMs) and advanced trajectory fusion techniques. We utilize the conventional scorers and the novel VLM-enhanced scorers. And we leverage a robust weight fusioner for quantitative aggregation and a powerful VLM-based fusioner for qualitative, context-aware decision-making. As the leading approach in the ICCV 2025 NAVSIM v2 End-to-End Driving Challenge, our SimpleVSF framework demonstrates state-of-the-art performance, achieving a superior balance between safety, comfort, and efficiency.

cs.RO↗
arXiv2025

Investigating The Smells of LLM Generated Code

Context: Large Language Models (LLMs) are increasingly being used to generate program code. Much research has been reported on the functional correctness of generated code, but there is far less on code quality. Objectives: In this study, we propose a scenario-based method of evaluating the quality of LLM-generated code to identify the weakest scenarios in which the quality of LLM generated code should be improved. Methods: The method measures code smells, an important indicator of code quality, and compares them with a baseline formed from reference solutions of professionally written code. The test dataset is divided into various subsets according to the topics of the code and complexity of the coding tasks to represent different scenarios of using LLMs for code generation. We will also present an automated test system for this purpose and report experiments with the Java programs generated in response to prompts given to four state-of-the-art LLMs: Gemini Pro, ChatGPT, Codex, and Falcon. Results: We find that LLM-generated code has a higher incidence of code smells compared to reference solutions. Falcon performed the least badly, with a smell increase of 42.28%, followed by Gemini Pro (62.07%), ChatGPT (65.05%) and finally Codex (84.97%). The average smell increase across all LLMs was 63.34%, comprising 73.35% for implementation smells and 21.42% for design smells. We also found that the increase in code smells is greater for more complex coding tasks and for more advanced topics, such as those involving object-orientated concepts. Conclusion: In terms of code smells, LLM's performances on various coding task complexities and topics are highly correlated to the quality of human written code in the corresponding scenarios. However, the quality of LLM generated code is noticeably poorer than human written code.

cs.SE↗
arXiv2025

Multi-Sensor Fusion for Extended Object Tracking Exploiting Active and Passive Radio Signals

Reliable and robust positioning of radio devices remains a challenging task due to multipath propagation, hardware impairments, and interference from other radio transmitters. A frequently overlooked but critical factor is the agent itself, e.g., the user carrying the device, which potentially obstructs line-of-sight (LOS) links to the base stations (anchors). This paper addresses the problem of accurate positioning in scenarios where LOS links are partially blocked by the agent. The agent is modeled as an extended object (EO) that scatters, attenuates, and blocks radio signals. We propose a Bayesian method that fuses ``active'' measurements (between device and anchors) with ``passive'' multistatic radar-type measurements (between anchors, reflected by the EO). To handle measurement origin uncertainty, we introduce an multi-sensor and multiple-measurement probabilistic data association (PDA) algorithm that jointly fuses all EO-related measurements. Furthermore, we develop an EO model tailored to agents such as human users, accounting for multiple reflections scattered off the body surface, and propose a simplified variant for low-complexity implementation. Evaluation on both synthetic and real radio measurements demonstrates that the proposed algorithm outperforms conventional PDA methods based on point target assumptions, particularly during and after obstructed line-of-sight (OLOS) conditions.

eess.SP↗
arXiv2025

FuXi-\beta: Towards a Lightweight and Fast Large-Scale Generative Recommendation Model

Scaling laws for autoregressive generative recommenders reveal potential for larger, more versatile systems but mean greater latency and training costs. To accelerate training and inference, we investigated the recent generative recommendation models HSTU and FuXi-$\alpha$, identifying two efficiency bottlenecks: the indexing operations in relative temporal attention bias and the computation of the query-key attention map. Additionally, we observed that relative attention bias in self-attention mechanisms can also serve as attention maps. Previous works like Synthesizer have shown that alternative forms of attention maps can achieve similar performance, naturally raising the question of whether some attention maps are redundant. Through empirical experiments, we discovered that using the query-key attention map might degrade the model's performance in recommendation tasks. To address these bottlenecks, we propose a new framework applicable to Transformer-like recommendation models. On one hand, we introduce Functional Relative Attention Bias, which avoids the time-consuming operations of the original relative attention bias, thereby accelerating the process. On the other hand, we remove the query-key attention map from the original self-attention layer and design a new Attention-Free Token Mixer module. Furthermore, by applying this framework to FuXi-$\alpha$, we introduce a new model, FuXi-$\beta$. Experiments across multiple datasets demonstrate that FuXi-$\beta$ outperforms previous state-of-the-art models and achieves significant acceleration compared to FuXi-$\alpha$, while also adhering to the scaling law. Notably, FuXi-$\beta$ shows an improvement of 27% to 47% in the NDCG@10 metric on large-scale industrial datasets compared to FuXi-$\alpha$. Our code is available in a public repository: https://github.com/USTC-StarTeam/FuXi-beta

cs.IR↗
arXiv2025

Unsupervised Image Super-Resolution Reconstruction Based on Real-World Degradation Patterns

The training of real-world super-resolution reconstruction models heavily relies on datasets that reflect real-world degradation patterns. Extracting and modeling degradation patterns for super-resolution reconstruction using only real-world low-resolution (LR) images remains a challenging task. When synthesizing datasets to simulate real-world degradation, relying solely on degradation extraction methods fails to capture both blur and diverse noise characteristics across varying LR distributions, as well as more implicit degradations such as color gamut shifts. Conversely, domain translation alone cannot accurately approximate real-world blur characteristics due to the significant degradation domain gap between synthetic and real data. To address these challenges, we propose a novel TripleGAN framework comprising two strategically designed components: The FirstGAN primarily focuses on narrowing the domain gap in blur characteristics, while the SecondGAN performs domain-specific translation to approximate target-domain blur properties and learn additional degradation patterns. The ThirdGAN is trained on pseudo-real data generated by the FirstGAN and SecondGAN to reconstruct real-world LR images. Extensive experiments on the RealSR and DRealSR datasets demonstrate that our method exhibits clear advantages in quantitative metrics while maintaining sharp reconstructions without over-smoothing artifacts. The proposed framework effectively learns real-world degradation patterns from LR observations and synthesizes aligned datasets with corresponding degradation characteristics, thereby enabling the trained network to achieve superior performance in reconstructing high-quality SR images from real-world LR inputs.

cs.CV↗
arXiv2025

Multi-Sensor Fusion of Active and Passive Measurements for Extended Object Tracking

This paper addresses the challenge of achieving robust and reliable positioning of a radio device carried by an agent, in scenarios where direct line-of-sight (LOS) radio links are obstructed by the agent. We propose a Bayesian estimation algorithm that integrates active measurements between the radio device and anchors with passive measurements in-between anchors reflecting off the agent. A geometry-based scattering measurement model is introduced for multi-sensor structures, and multiple object-related measurements are incorporated to formulate an extended object probabilistic data association (PDA) algorithm, where the agent that blocks, scatters and attenuates radio signals is modeled as an extended object (EO). The proposed approach significantly improves the accuracy during and after obstructed LOS conditions, outperforming the conventional PDA (which is based on the point-target-assumption) and methods relying solely on active measurements.

eess.SP↗