Searcharxiv⌕ Search

arXiv subjects

Zifan Zhang

Publications and source records attributed to Zifan Zhang.

At least 19 recordsLinked to original sources

GPRM: Global Perspective Process Reward Model

Process Reward Models (PRMs) are widely used for step-level evaluation of multi-step reasoning in large language models, but existing methods are often limited by weak historical dependency modeling and insufficient global awareness. They typically fail to leverage prior step evaluations for coherent judgment and lack visibility into subsequent reasoning, leading to suboptimal assessment of intermediate steps. To address this issue, we propose the Global Perspective Process Reward Model (GPRM), which jointly models the problem, previous reasoning steps and their evaluations, the current step, and future steps to enable history-aware and globally informed process evaluation. We further introduce a structured evaluation framework along four dimensions---look back, look ahead, self-check, and goal alignment---to assess prior consistency, future plausibility, step validity, and task alignment. Based on this framework, we construct datasets for supervised fine-tuning (SFT) and reinforcement learning (GRPO). Experiments show that, despite using less training data and smaller model sizes, GPRM outperforms strong baseline models by 7.8 and 5.4 points on PRMBench and ProcessBench, respectively. GPRM also generalizes well across tasks, surpassing the GPT-4.1 baseline on Agent Error Bench, improving strong baselines by around 4 points under test-time scaling (TTS), and achieving competitive performance on PPE.

cs.CL↗

MedPRMBench: A Fine-grained Benchmark for Process Reward Models in Medical Reasoning

Medical reasoning can reach a plausible answer while overlooking an unsafe or unsupported intermediate step. Process-Level Reward Models (PRMs) can verify such steps, yet existing PRM benchmarks largely target mathematics rather than patient-specific clinical errors. We introduce \textbf{MedPRMBench}, a step-level medical reasoning benchmark with 14 clinical error types, including five new and two clinically re-scoped types, and rule-defined severity grades. Built from seven medical QA sources via structured error construction and review by 86 physicians from Grade-A tertiary hospitals, it provides 6{,}500 evaluation questions, 13{,}000 reasoning chains, and 113{,}910 step labels, plus a leakage-clean 6{,}870-question training split. The best of 20 prompted critics reaches only 75.4\% PRMScore, and critics differ sharply in error-type profiles and operating points. A Qwen3-8B PRM trained on the split reaches 86.6\% PRMScore and 89.0\% strict first-error localization; in matched ablations, physician review raises error recall by 6.55 points and cuts clean-chain false alarms from 5.6\% to 2.1\%. The training signal replicates across model families; without further training, the PRM transfers to ProcessBench and PRMBench, exceeding six of nine math-trained PRMs on each.

cs.CL↗

Efficient Hierarchical Transformers for Representing Log Data

Transformers have gained widespread acclaim for their versatility in handling diverse data structures, yet their application to log data remains underexplored. Log data, characterized by its hierarchical, dictionary-like structure, poses unique challenges when processed using conventional transformer models. Traditional methods often rely on manually crafted templates for parsing logs, a process that is labor-intensive and lacks generalizability. Additionally, the linear treatment of log sequences by standard transformers neglects the rich, nested relationships within log entries, leading to suboptimal representations and excessive memory usage. To address these issues, we introduce HLogformer, a novel hierarchical transformer framework specifically designed for log data. HLogformer leverages the hierarchical structure of log entries to significantly reduce memory costs and enhance representation learning. Unlike traditional models that treat log data as flat sequences, our framework processes log entries in a manner that respects their inherent hierarchical organization. This approach ensures comprehensive encoding of both fine-grained details and broader contextual relationships. Our contributions are threefold: First, HLogformer is the first framework to design a dynamic hierarchical transformer tailored for dictionary-like log data. Second, it dramatically reduces memory costs associated with processing extensive log sequences. Third, comprehensive experiments demonstrate that HLogformer more effectively encodes hierarchical contextual information, proving to be highly effective for downstream tasks such as synthetic anomaly detection and product recommendation.

cs.LG↗

Network-in-the-Loop at Scale: GPU-Batched 5G Simulation for Massively Parallel Robot Learning

Massively parallel GPU simulators train multi-robot policies in thousands of environments, and many fleets use private Fifth-Generation (5G) networks, where each robot's delay depends on its teammates' traffic. Network-in-the-loop training places a simulated 5G network inside this loop. However, GPU robot simulators reduce the network to an independent delay per message, while packet-level simulators run one scenario per CPU process and cannot keep pace with thousands of parallel environments. To bridge this gap, we present Isaac-Net, a GPU-batched 5G New Radio (NR) module that advances the uplink of thousands of environments in lockstep with Isaac Lab physics. Isaac-Net simulates every slot, the 0.5~ms interval in which the base station decides which robots transmit, for all environments at once. Extensive experiments confirm that its NR engine reproduces the median delay of ns-3 5G-LENA across loads, with a median delay 5--10\% low on an unseen carrier and 9\% high at 32 robots per environment in closed loop. The engine also reproduces the Age of Information (AoI), the age of each robot's newest delivered report, while an independent delay per message leaves the AoI tail about three times too light. In a configuration validated against 5G-LENA, Isaac-Net keeps the network in the loop for about one million robots on one GPU at 83\% of the Isaac Lab rate without the network, measured under a random policy. Isaac-Net is open source at https://github.com/ZzZTripleZzZ/isaac-net

cs.NI↗

OpenTwin: Closed-Loop Digital Twins for Trustworthy Policy Deployment in Open RAN

In open radio access networks (O-RAN), the near-real-time RAN Intelligent Controller (RIC) hosts third-party xApps whose training and validation risk disrupting the operational network. Disaggregation amplifies this risk, as no single party can certify a control action end to end. Indeed, our testbed shows an E2 control request reported as successful while the base station never applies the change. Digital twins (DTs) promise safe policy evaluation, yet existing O-RAN DTs largely rely on hand-crafted models, run without feedback from the deployment, and never say how often their predictions can be trusted. To fill this gap, we present OpenTwin, a closed-loop framework that learns the simulator configuration reproducing an operating deployment streamed measurements, certifies the resulting DT by re-simulation, calibrates it online, and evaluates each xApp action before it executes on the physical network. Every trust decision carries an error rate bounded by an operator-prescribed budget, with a drift detector limiting needless resynchronizations and a conformal fidelity gate admitting an action only when its predicted outcome range lies in the safe region. Extensive experiments across simulation and real-world testbeds confirm single-digit percentage error in reproduced measurements, false approvals an order of magnitude below every budget from 0.05 to 0.30, and a gated energy-saving xApp that retains roughly 40% of the saving achievable with perfect foresight.

cs.NI↗

Giant bulk photovoltaic effect driven by interfacial symmetry breaking in MoS2/Ta2NiSe5 heterostructures

Van der Waals (vdW) heterostructures offer a versatile platform for engineering unconventional bulk photovoltaic (BPV) effect through interfacial symmetry breaking. However, the coexistence of multiple photophysical mechanisms, driven by structural complexity, spontaneous charge transfer, and strong interlayer coupling, often obscures the microscopic origin of the BPV response and hinders its rational optimization. Here, we demonstrate a pronounced BPV effect localized at the overlap region of a cross-bar MoS2/Ta2NiSe5 vdW heterostructure, where symmetry breaking induced by vertical stacking lifts the inversion center of MoS2. The orthogonal device geometry enables the independent probing of intralayer and interfacial photoresponse pathways, facilitating clear separation of competing mechanisms. Spontaneous interfacial charge transfer between MoS2 and Ta2NiSe5 further establishes a strong interlayer electronic coupling. By modulating the interlayer potential landscape through gate voltage and vertical electric fields, we achieve an optimized zero-bias photocurrent density of 247 A/cm2 and a BPV coefficient of 0.99 V-1. Supported by theoretical modelling, our results illustrate how minimalist device geometry can transform complex heterostructures into experimentally tractable platforms. This strategy paves the way for analyzing and optimizing interface-driven BPV effect, with implications for self-powered optoelectronics, broadband photodetection, and energy-harvesting nanodevices.

cond-mat.mes-hall↗

CENTILE: A Telemetry Foundation Model Evaluated by the Decisions It Drives

Modern computing and networking infrastructure emits telemetry continuously, yet operators convert it into decisions with a separate predictor per task, entity, and horizon. One generative model, pretrained once over an operator's own event streams, could replace this fleet, an approach that already scales to high-cardinality streams in recommendation systems. However, point-forecast error on operational telemetry saturates near simple last-value baselines, so lower error alone need not improve the decisions it feeds. To close this gap, we present \sys, a generative foundation model for network and systems telemetry, evaluated by replaying the decisions its calibrated conditional quantiles drive. \sys treats heterogeneous telemetry as event-driven, irregularly timed entity streams and serves flexible forecast horizons in a single pass, requiring no future timestamps. To our knowledge, \sys is the first pretrained telemetry model to improve both HPC scheduling and network provisioning decisions under replay, its runtime estimator transferring zero-shot across months and its pretrained weights across domains from hours of target data. Extensive experiments on HPC job logs and network traffic confirm that \sys lowers the mean bounded slowdown of backfilling by up to approximately $77\%$ over deployed user estimates and roughly halves the deployed rule's violation rate. Our code is available at https://github.com/ZzZTripleZzZ/all-in-one.

cs.NI↗

Policy of Thoughts: Scaling Test-Time Training for LLM Reasoning via Online Policy Evolution

Large language models (LLMs) struggle with complex, long-horizon reasoning due to instability caused by their frozen policy assumption. Current test-time scaling methods treat execution feedback merely as an external signal for filtering or rewriting trajectories, without internalizing it to improve the underlying reasoning strategy. Inspired by Popper's epistemology of "conjectures and refutations," we argue that intelligence requires real-time evolution of the model's policy through learning from failed attempts. We introduce Policy of Thoughts (PoT), a framework that recasts reasoning as a within-instance online optimization process. PoT first generates diverse candidate solutions via an efficient exploration mechanism, then uses Group Relative Policy Optimization (GRPO) to update a transient LoRA adapter based on execution feedback. This closed-loop design enables dynamic, instance-specific refinement of the model's reasoning priors. Experiments show that PoT dramatically boosts performance: a 4B model achieves 49.71% accuracy on LiveCodeBench, outperforming GPT-4o and DeepSeek-V3 despite being over 50 smaller.

cs.AI↗

Securing Autonomous Vehicle Systems via Twin-Aware Federated Reinforcement Learning

Federated reinforcement learning (FRL) is crucial for enabling collaborative learning across multiple agents without sharing raw data, thereby enhancing privacy and scalability in the decision-making process within dynamic vehicular environments. However, poisoning attacks pose a significant threat to the security and reliability of FRL-based systems, particularly in safety-critical autonomous driving, where this vulnerability remains largely unexplored. These attacks can compromise the global control model by subtly injecting malicious system parameters, leading to potential hazards. To counter these challenges, we present \alg (\underline{Sec}ure \underline{A}ggregation with \underline{p}oisoning-\underline{p}revention and historical reinforcement) as a defensive framework aimed at enhancing the robustness of FRL systems designed for safety-critical driving scenarios. \alg strategically integrates digital twins for rehearsal-based learning and leverages historical aggregated model parameters along with a selected central gradient to ensure that only benign data is aggregated, effectively mitigating the influence of malicious agents. Theoretical guarantees are provided for the convergence performance of \alg in the presence of poisoning attacks. We also validate the effectiveness of \alg using developed digital twins that model realistic highway environments to evaluate the control of autonomous vehicles under adversarial conditions.

cs.CR↗

Customized Generative AI Agent for Transportation Engineering Practice: A Development and Continued Pre-training Guideline

Recent advancements in generative artificial intelligence (AI) and large language models (LLMs) have shown significant promise in automating complex reasoning, summarization, and question-answering tasks. However, the effectiveness of general-purpose LLMs in specialized engineering domains remains limited due to insufficient exposure to technical standards, engineering terminology, and domain-specific semantics. This study proposes a systematic approach to developing a customized generative AI agent for transportation engineering applications. A curated corpus of U.S. transportation manuals, design guidelines, and regulatory documents is used to conduct continued pretraining of six state-of-the-art LLMs through a unified low-rank adaptation (LoRA) framework. The training process is monitored to ensure convergence and model stability. Performance is evaluated using standard natural language processing metrics, including BLEU-4 and ROUGE, with Qwen2.5-7B and LLaMA-3.1-8B demonstrating the highest domain alignment and response quality. Results validate the effectiveness of LoRA-based adaptation in improving LLM performance on technical content interpretation and context-specific reasoning. This work contributes a reproducible development framework for constructing domain-specialized generative AI agents, supporting broader deployment in transportation research, design, planning, and policy analysis.

cs.AI↗

"Refactoring Runaway": Understanding and Mitigating Tangled Refactorings in Coding Agents for Issue Resolution

Recent advances in coding agents have shown remarkable progress in software issue resolution. In practice, real-world issues are typically bug fixes or feature requests in which human developers naturally incorporate refactoring as part of the resolution process, resulting in tangled refactoring. Since LLMs are trained on large-scale open-source repositories, coding agents may inherit such behaviors. In this paper, we conduct an empirical study on Multi-SWE-bench, analyzing 3,691 valid patches generated by three agent frameworks with 12 LLMs. We find that coding agents introduce tangled refactorings less frequently (21.43% vs. 36.72%) and with lower intensity (0.66 vs. 1.75) than human developers, although they exhibit a broader diversity of refactoring types. Logistic regression analysis further shows that tangled refactorings are strongly associated with reduced compilability, while exhibiting no significant association with functional correctness. Based on these findings, we propose a refactoring-aware refinement approach that assesses the necessity and safety of tangled refactorings and selectively removes or repairs problematic operations. Our approach improves compilability from 19.34% to 38.33%, and additionally resolves 2.79% previously unresolved issues. Overall, this work presents the first step towards understanding tangled refactoring practices in agentic issue resolution and opens up avenues for future work.

cs.SE↗

Network Digital Untwinning: Towards Backward Optimization of Digital Twins

Network digital twins (NDTs) are transforming network management by offering precise virtual replicas of physical network systems. However, their reliance on diverse and sensitive data introduces significant challenges related to data management, regulatory compliance, and user privacy. In scenarios where selective data removal is necessary, such as device deactivation, network reconfiguration, or regulatory compliance, traditional approaches often fall short of preserving the integrity of the twin model. To address this gap, we introduce a network digital untwinning framework that enables the targeted removal of deprecated NDT contributions while maintaining model integrity. Our approach comprises two complementary components: Single Request Untwinning (\algO) and Parallel Request Untwinning (\algM) mechanisms. \algO leverages connectivity metrics based on geographical proximity, data distribution, and network-level attributes to identify and remove the target NDT along with its propagating influence. This is achieved through an optimally selected rollback checkpoint augmented with injected Gaussian noise, followed by a precise remapping phase. \algM extends this mechanism to efficiently handle multiple removal requests by clustering NDTs with similar attributes and performing a coordinated rollback and untwinning schedule. We provide theoretical guarantees on model indistinguishability from scratch-built twins, and validate the framework through extensive experiments on real-world traffic data, demonstrating its effectiveness and operational efficiency.

cs.NI↗

INSIGHT: Enhancing Autonomous Driving Safety through Vision-Language Models on Context-Aware Hazard Detection and Edge Case Evaluation

Autonomous driving systems face significant challenges in handling unpredictable edge-case scenarios, such as adversarial pedestrian movements, dangerous vehicle maneuvers, and sudden environmental changes. Current end-to-end driving models struggle with generalization to these rare events due to limitations in traditional detection and prediction approaches. To address this, we propose INSIGHT (Integration of Semantic and Visual Inputs for Generalized Hazard Tracking), a hierarchical vision-language model (VLM) framework designed to enhance hazard detection and edge-case evaluation. By using multimodal data fusion, our approach integrates semantic and visual representations, enabling precise interpretation of driving scenarios and accurate forecasting of potential dangers. Through supervised fine-tuning of VLMs, we optimize spatial hazard localization using attention-based mechanisms and coordinate regression techniques. Experimental results on the BDD100K dataset demonstrate a substantial improvement in hazard prediction straightforwardness and accuracy over existing models, achieving a notable increase in generalization performance. This advancement enhances the robustness and safety of autonomous driving systems, ensuring improved situational awareness and potential decision-making in complex real-world scenarios.

cs.CV↗

HyTRec: A Hybrid Temporal-Aware Attention Architecture for Long Behavior Sequential Recommendation

Modeling long sequences of user behaviors has emerged as a critical frontier in generative recommendation. However, existing solutions face a dilemma: linear attention mechanisms achieve efficiency at the cost of retrieval precision due to limited state capacity, while softmax attention suffers from prohibitive computational overhead. To address this challenge, we propose HyTRec, a model featuring a Hybrid Attention architecture that explicitly decouples long-term stable preferences from short-term intent spikes. By assigning massive historical sequences to a linear attention branch and reserving a specialized softmax attention branch for recent interactions, our approach restores precise retrieval capabilities within industrial-scale contexts involving ten thousand interactions. To mitigate the lag in capturing rapid interest drifts within the linear layers, we furthermore design Temporal-Aware Delta Network (TADN) to dynamically upweight fresh behavioral signals while effectively suppressing historical noise. Empirical results on industrial-scale datasets confirm the superiority that our model maintains linear inference speed and outperforms strong baselines, notably delivering over 8% improvement in Hit Rate for users with ultra-long sequences with great efficiency.

cs.IR↗

Credit Where It is Due: Cross-Modality Connectivity Drives Precise Reinforcement Learning for MLLM Reasoning

Reinforcement Learning with Verifiable Rewards (RLVR) has significantly advanced the reasoning capabilities of Multimodal Large Language Models (MLLMs), yet how visual evidence is integrated during reasoning remains poorly understood. We explore multimodal RLVR through the lens of cross-modal attention connectivity and find that only a small fraction of tokens (approximately 15%) exhibit strong visual-textual coupling. These high-connectivity tokens act as anchors that ground reasoning in the image, while the majority follow linguistic patterns. During RLVR training, credit assignment naturally concentrates on these anchors, sharpening their visual grounding over time. Building on this insight, we propose Anchor-Token Reinforcement Learning (AT-RL), a lightweight framework that selectively reinforces high-connectivity tokens via graph-based clustering of attention topology. Evaluated across the series (3B-32B), AT-RL introduces only 1.2% overhead yet enables the 32B model to surpass the 72B-Instruct baseline on MathVista (80.2), with consistent gains observed across STEM, video and general tasks. Conversely, training solely on low-connectivity tokens causes severe degradation, confirming that effective multimodal RL hinges on precise credit assignment to visual anchors. Our work reveals that reasoning quality is governed not by token quantity but by the fidelity of cross-modal anchoring.

cs.AI↗

Agentic Proposing: Enhancing Large Language Model Reasoning via Compositional Skill Synthesis

Advancing complex reasoning in large language models relies on high-quality, verifiable datasets, yet human annotation remains cost-prohibitive and difficult to scale. Current synthesis paradigms often face a recurring trade-off: maintaining structural validity typically restricts problem complexity, while relaxing constraints to increase difficulty frequently leads to inconsistent or unsolvable instances. To address this, we propose Agentic Proposing, a framework that models problem synthesis as a goal-driven sequential decision process where a specialized agent dynamically selects and composes modular reasoning skills. Through an iterative workflow of internal reflection and tool-use, we develop the Agentic-Proposer-4B using Multi-Granularity Policy Optimization (MGPO) to generate high-precision, verifiable training trajectories across mathematics, coding, and science. Empirical results demonstrate that downstream solvers trained on agent-synthesized data significantly outperform leading baselines and exhibit robust cross-domain generalization. Notably, a 30B solver trained on only 11,000 synthesized trajectories achieves a state-of-the-art 91.6% accuracy on AIME25, rivaling frontier-scale proprietary models such as GPT-5 and proving that a small volume of high-quality synthetic signals can effectively substitute for massive human-curated datasets.

cs.AI↗

Socratic-Geo: Synthetic Data Generation and Geometric Reasoning via Multi-Agent Interaction

Multimodal Large Language Models (MLLMs) have significantly advanced vision-language understanding. However, even state-of-the-art models struggle with geometric reasoning, revealing a critical bottleneck: the extreme scarcity of high-quality image-text pairs. Human annotation is prohibitively expensive, while automated methods fail to ensure fidelity and training effectiveness. Existing approaches either passively adapt to available images or employ inefficient random exploration with filtering, decoupling generation from learning needs. We propose Socratic-Geo, a fully autonomous framework that dynamically couples data synthesis with model learning through multi-agent interaction. The Teacher agent generates parameterized Python scripts with reflective feedback (Reflect for solvability, RePI for visual validity), ensuring image-text pair purity. The Solver agent optimizes reasoning through preference learning, with failure paths guiding Teacher's targeted augmentation. Independently, the Generator learns image generation capabilities on accumulated "image-code-instruction" triplets, distilling programmatic drawing intelligence into visual generation. Starting from only 108 seed problems, Socratic-Solver achieves 49.11 on six benchmarks using one-quarter of baseline data, surpassing strong baselines by 2.43 points. Socratic-Generator achieves 42.4% on GenExam, establishing new state-of-the-art for open-source models, surpassing Seedream-4.0 (39.8%) and approaching Gemini-2.5-Flash-Image (43.1%).

cs.CV↗

Socratic-Zero : Bootstrapping Reasoning via Data-Free Agent Co-evolution

Recent breakthroughs in large language models (LLMs) on reasoning tasks rely heavily on massive, high-quality datasets-typically human-annotated and thus difficult to scale. While data synthesis or distillation offers a promising alternative, existing methods struggle with inconsistent data quality and an inability to dynamically adapt to the evolving capabilities of the model, leading to suboptimal training signals. To address these limitations, we introduce Socratic-Zero, a fully autonomous framework that generates high-quality training data from minimal seed examples through the co-evolution of three agents: the Teacher, the Solver, and the Generator. The Solver continuously refines its reasoning by learning from preference feedback on both successful and failed trajectories; the Teacher adaptively crafts increasingly challenging questions based on the Solver's weaknesses; and the Generator distills the Teacher's question-design strategy to enable scalable, high-fidelity curriculum generation. This closed-loop system produces a self-improving curriculum-requiring no pre-existing tasks or labels. Remarkably, starting from only 100 seed questions, our Socratic-Solver-8B achieves an average gain of +20.2 percentage points over prior data synthesis methods across seven mathematical reasoning benchmarks (AMC23, AIME24-25, Olympiad, MATH-500, Minerva, and GSM8K), with consistent gains on both Qwen3 and GLM4 series models. Even more surprisingly, synthetic data from Socratic-Generator-32B enables student LLMs to achieve superior performance compared to other state-of-the-art (SOTA) commercial LLMs on these benchmarks, including Qwen3-235B-A22B, DeepSeek-V3.1-671B, GPT-5, Gemini-2.5-Pro, Grok-4, and Claude-4.1-Opus.

cs.CL↗