arXiv subjects

Xinyu Liu

Publications and source records attributed to Xinyu Liu.

At least 91 records · Page 5Linked to original sources

arXiv2026

Depletion-mode N-polar AlN-based high electron mobility transistors with improved on/off ratios

We report N-polar AlN-based high-electron mobility transistors (HEMTs) with a GaN channel thickness of 5.2 nm on N-polar AlN on sapphire. The threshold voltage is around -2.4 to -3.0 V with saturation currents over 240 mA/mm and on/off ratios as high as 10,000, much higher than previously reported N-polar AlN-based HEMTs. The high on/off ratio is attributed to the use of an abrupt AlN/GaN heterostructure with a dedicated AlN transition layer, together with improved gate leakage. The high frequency properties as well as the on-resistance of ~20 Ohm mm are all limited by the 2000 Ohm/square sheet resistance of the channel layer.

cond-mat.mtrl-sci↗

arXiv2026

Exploiting LLM Agent Supply Chains via Payload-less Skills

Autonomous agents powered by Large Language Models (LLMs) acquire external functionalities through third-party skills available in open marketplaces. Adopting these integrations broadens the potential attack surface, prompting a need for systematic security evaluation. Current auditing mechanisms are effective at identifying explicit code payloads and predefined threat contents through security scanning. These detection mechanisms are bypassed if malicious behaviors lack direct injection and are instead synthesized dynamically at runtime through the agent's inherent generative capabilities. Exploring this blind spot, we introduce Semantic Compliance Hijacking (SCH), a payload-less supply chain attack targeting autonomous coding environments. The SCH approach translates malicious goals into unstructured natural language instructions formatted as necessary compliance rules, leading the agent to generate and execute unauthorized code. To assess the real-world viability of this attack, we developed an automated pipeline to evaluate its effectiveness across a test matrix comprising three mainstream agent frameworks and three distinct foundation models using contextualized scenarios. The findings demonstrate the pervasive nature of this threat, with SCH achieving peak success rates of up to 77.67% for confidentiality breaches and 67.33% for Remote Code Execution (RCE) under the most vulnerable configurations. Furthermore, the introduction of Multi-Skill Automated Optimization (MS-AO) further boosted attack efficacy. By omitting recognizable Abstract Syntax Tree (AST) signatures and explicit harmful intents, the manipulated skill files maintained a 0.00% detection rate, evading current scanning tools. This research highlights an underexplored attack surface within agent supply chains, pointing to a necessary transition from signature-based detection models toward semantic intent validation.

cs.CR↗

arXiv2026

COPRA: Conditional Parameter Adaptation with Reinforcement Learning for Video Anomaly Detection

Vision-language models (VLMs) have shown strong performance in video anomaly detection (VAD) while providing interpretable predictions. However, existing VLM-based VAD methods suffer from a fundamental mismatch between training and inference in both data distribution and model configuration. First, most approaches rely on static post-training adaptation, limiting generalization under distribution shifts such as unseen environments or anomaly types. Second, they train VLMs on sparse frames from long videos, but perform inference on densely sampled short segments, creating inconsistencies between training and testing. To address these limitations, we propose COPRA, a conditional parameter adaptation framework for VLM-based VAD. Instead of fixed prompts or shared parameter updates, COPRA generates input-specific parameter updates to dynamically adapt a frozen VLM for each video segment during both training and inference. Experiments show strong performance on standard VAD benchmarks, consistently outperforming static baselines in both in-domain and cross-domain settings. Moreover, COPRA generalizes beyond VAD to unseen tasks such as multiple-choice Video Question Answering and Dense Captioning. These results highlight COPRA as an effective weight-space generation framework for scalable, adaptive, and context-aware video understanding. The code will be released at https://github.com/THE-MALT-LAB/COPRA

cs.CV↗

arXiv2026

MathlibPR: Pull Request Merge-Readiness Benchmark for Formal Mathematical Libraries

The ecosystem of Lean and Mathlib has become the de facto standard for large language model (LLM) assisted formal reasoning with remarkable successes in recent years. Those successes, however, only consume Mathlib as an essential dependency but do not directly contribute to it. In the meantime, the growth of Mathlib has recently been bottlenecked by the review process, which requires human reviewers to judge whether proposed pull requests (PRs) follow the Mathlib's conventions and are worth integrating as part of a shared mathematical infrastructure. This leads to our central question: can LLMs help review Mathlib PRs? To this end, we introduce MathlibPR, a benchmark built from real Mathlib4 PR histories. We further propose a staged evaluation protocol and use it to evaluate both LLM models (e.g., DeepSeek, Qwen, Goedel, and Kimina) and LLM agents (e.g., Codex and Claude Code). Surprisingly, both LLM models and LLM agents struggle to distinguish merge-ready PRs from build-passing PRs that were revised or never merged. By turning Mathlib PR histories into a supervised signal, MathlibPR provides a step toward reviewer assistants and reward models that could help evaluate PRs and steer LLMs toward producing merge-ready Mathlib contributions.

cs.LO↗

arXiv2026

Offline Two-Player Zero-Sum Markov Games with KL Regularization

We study the problem of learning Nash equilibria in offline two-player zero-sum Markov games. While existing approaches often rely on explicit pessimism to address distribution shift, we show that KL regularization alone suffices to stabilize learning and guarantee convergence. We first introduce Regularized Offline Sequential Equilibrium (ROSE), a theoretical framework that achieves a fast $\widetilde{\mathcal{O}}(1/n)$ convergence rate under \textit{unilateral concentrability}, improving over the standard $\widetilde{\mathcal{O}}(1/\sqrt{n})$ rates in unregularized settings. We then propose Sequential Offline Self-play Mirror Descent (SOS-MD), a practical model-free algorithm based on least-squares value estimation and iterative self-play updates. We prove that the last iterate of SOS-MD attains the same $\widetilde{\mathcal{O}}(1/n)$ statistical rate up to a vanishing optimization error of order $\widetilde{\mathcal{O}}(1/\sqrt{T})$ in the number of self-play iterations $T$.

cs.LG↗

arXiv2026

AIvaluateXR: An Evaluation Framework for on-Device AI in XR with Benchmarking Results

The deployment of large language models (LLMs) on extended reality (XR) devices has great potential to advance the field of human-AI interaction. In the case of direct, on-device model inference, selecting the appropriate model and device for specific tasks remains challenging. In this paper, we present AIvaluateXR, a comprehensive evaluation framework for benchmarking LLMs running on XR devices. To demonstrate the framework, we deploy 17 selected LLMs across four XR platforms: Magic Leap 2, Meta Quest 3, Vivo X100s Pro, and Apple Vision Pro, and conduct an extensive evaluation. Our experimental setup measures four key metrics: performance consistency, processing speed, memory usage, and battery consumption. For each of the 68 model-device pairs, we assess performance under varying string lengths, batch sizes, and thread counts, analyzing the trade-offs for real-time XR applications. We propose a unified evaluation method based on the 3D Pareto Optimality theory to select the optimal device-model pairs from quality and speed objectives. Additionally, we compare the efficiency of on-device LLMs with client-server and cloud-based setups, and evaluate their accuracy on two interactive tasks. We believe our findings offer valuable insight to guide future optimization efforts for LLM deployment on XR devices. Our evaluation method can be used as standard groundwork for further research and development in this emerging field. The source code and supplementary materials are available at: www.nanovis.org/AIvaluateXR.html

cs.DC↗

arXiv2026

GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents

We present GLM-5V-Turbo, a step toward native foundation models for multimodal agents. As foundation models are increasingly deployed in real environments, agentic capability depends not only on language reasoning, but also on the ability to perceive, interpret, and act over heterogeneous contexts such as images, videos, webpages, documents, GUIs. GLM-5V-Turbo is built around this objective: multimodal perception is integrated as a core component of reasoning, planning, tool use, and execution, rather than as an auxiliary interface to a language model. This report summarizes the main improvements behind GLM-5V-Turbo across model design, multimodal training, reinforcement learning, toolchain expansion, and integration with agent frameworks. These developments lead to strong performance in multimodal coding, visual tool use, and framework-based agentic tasks, while preserving competitive text-only coding capability. More importantly, our development process offers practical insights for building multimodal agents, highlighting the central role of multimodal perception, hierarchical optimization, and reliable end-to-end verification.

cs.CV↗

arXiv2026

Measurement-Based Quantum Diffusion Models

We introduce measurement-based quantum diffusion models that bridge classical and quantum diffusion theory through randomized weak measurements. The measurement-based approach naturally generates stochastic quantum trajectories while preserving purity at the trajectory level and inducing depolarization at the ensemble level. We address two quantum state generation problems: trajectory-level recovery of pure state ensembles and ensemble-average recovery of mixed states. For trajectory-level recovery, we establish that quantum score matching is mathematically equivalent to learning unitary generators for the reverse process. For ensemble-average recovery, we introduce local Petz recovery maps for states with finite correlation length and classical shadow reconstruction for general states, both with rigorous error bounds. Our framework establishes Petz recovery maps as quantum generalizations of reverse Fokker-Planck equations, providing a rigorous bridge between quantum recovery channels and classical stochastic reversals. This work enables new approaches to quantum state generation with potential applications in quantum information science.

quant-ph↗

arXiv2026

Almost Sure Convergence Rates of Stochastic Approximation and Reinforcement Learning via a Poisson-Moreau Drift

Establishing almost sure convergence rates for stochastic approximation and reinforcement learning under Markovian noise is a fundamental theoretical challenge. We make progress towards this challenge for a class of stochastic approximation algorithms whose expected updates are contractive, a setting that arises in many reinforcement learning algorithms such as $Q$-learning and linear temporal difference learning. Specifically, for a power-law learning rate $O(n^{-η})$ with $η\in (1/2, 1)$, we obtain an almost sure convergence rate arbitrarily close to $o(n^{1 - 2η})$. For a harmonic learning rate $O(n^{-1})$, we obtain an almost sure convergence rate arbitrarily close to $o(n^{-1})$, which we argue is a strong result because it is close to the optimal rate $O(n^{-1}\log\log n)$ given by the law of the iterated logarithm (for a special case of i.i.d. noise). Key to our analysis is a novel Lyapunov drift construction that applies a Poisson-equation based correction for Markovian noise to the well-established Moreau-envelope smoothing for the contractive mapping.

cs.LG↗

arXiv2026

Convergence and Emergence of In-Context Reinforcement Learning with Chain of Thought

In-context reinforcement learning (ICRL) refers to the ability of RL agents to adapt to new tasks at inference time without parameter updates by conditioning on additional context. Recent empirical studies further demonstrate that Chain-of-Thought (CoT) generation can amplify this ICRL capability. This paper is the first to provide a theoretical understanding on how CoT interacts with ICRL. We conduct our analysis in a policy evaluation setup with linear Transformer. We prove that with specific Transformer parameters, the CoT generation process is equivalent to repeatedly executing temporal difference learning updates. Additionally, we provide finite sample convergence analysis showing that the policy evaluation error decreases geometrically with CoT length and eventually saturates at a statistical floor determined by the context length. We also prove that the desired Transformer parameters are a global minimizer of the pretraining loss, providing a theoretical understanding on the empirical emergence of those parameters.

cs.LG↗

arXiv2026

OracleTSC: Oracle-Informed Reward Hurdle and Uncertainty Regularization for Traffic Signal Control

Transparent decision-making is essential for traffic signal control (TSC) systems to earn public trust. However, traditional reinforcement learning-based TSC methods function as black boxes with limited interpretability. Although large language models (LLMs) can provide natural language reasoning, reinforcement finetuning for TSC remains unstable because feedback is sparse and delayed, while most actions produce only marginal changes in congestion metrics. We introduce OracleTSC, which stabilizes LLM-based TSC through two mechanisms: (1) a reward hurdle mechanism that filters weak learning signals by subtracting a calibrated threshold from environmental rewards, and (2) uncertainty regularization that maximizes the probability of the selected response to encourage consistent decisions across sampled outputs. Experiments on the LibSignal benchmark show that OracleTSC enables a compact LLaMA3-8B model to substantially improve traffic efficiency, achieving a 75% reduction in travel time and a 67% decrease in queue length compared with the pretrained baseline while preserving interpretability through natural language explanations. OracleTSC also demonstrates strong cross-intersection generalization: a policy trained on one intersection transfers to a structurally different intersection with 17% lower travel time and 39% lower queue length without additional finetuning. These results suggest that uncertainty-aware reward shaping can improve the stability and effectiveness of reinforcement fine-tuning for TSC.

cs.AI↗

arXiv2026

SiblingRepair: Sibling-Based Multi-Hunk Repair with Large Language Models

Developers often make similar mistakes across code locations implementing related functionalities. These locations, called siblings, share similar issues and require similar fixes. Accurately identifying siblings and consistently repairing them are crucial for automated program repair. Hercules is a SOTA technique designed for sibling repair. However, it is limited by strong assumptions about sibling locations and commit-history availability, rigid AST-based sibling matching, and inflexible template-based patch generation. To address these limitations, we present SiblingRepair, a new LLM-based multi-hunk APR technique specialized for sibling repair. Starting from a suspicious location identified by spectrum-based fault localization, SiblingRepair searches for semantically related sibling candidates using token- and embedding-based code matching, without restricting discovery to failing-test coverage or commit history. It then uses an LLM to identify failure-relevant siblings and generate consistent patches through two complementary strategies: simultaneous repair, which jointly repairs siblings, and iterative repair, which progressively analyzes candidates for patch construction. SiblingRepair further preserves promising patches generated from earlier suspicious locations and combines them into generalized multi-hunk patches. We evaluate SiblingRepair on the Defects4J and GHRB benchmarks. The results show that SiblingRepair substantially outperforms SOTA multi-hunk repair techniques including Hercules. Our evaluation further demonstrates its repair efficiency, the effectiveness of its sibling detection and repair components, and limited impact of the LLM data leakage on the results. Overall, SiblingRepair advances automated sibling and general multi-hunk repair.

cs.SE↗

arXiv2026

AesRM: Improving Video Aesthetics with Expert-Level Feedback

Despite rapid advances in photorealistic video generation, real-world applications such as filmmaking require video aesthetics, e.g., harmonious colors and cinematic lighting, beyond visual fidelity. Prior work on visual aesthetics largely focuses on images, often reducing aesthetics to coarse definitions, e.g., visual pleasure, without a rigorous and systematic evaluation. To improve video aesthetics, we propose a hierarchical rubric that decomposes video aesthetics into three core dimensions, Visual Aesthetics (VA), Visual Fidelity (VF), and Visual Plausibility (VP), with 15 fine-grained criteria, e.g., shot composition. This framework enables a large-scale expert-annotated preference dataset and an evaluation benchmark, AesVideo-Bench, containing about 2500 video pairs with expert annotations on VA, VF, and VP. We then build a family of Video Aesthetic Reward Models (AesRM): AesRM-Base, which directly predicts pairwise preferences on these dimensions to provide efficient post-training rewards, and AesRM-CoT, which additionally generates CoT aligned with all 15 criteria to improve assessment interpretability. Specifically, we train AesRM with a three-stage progressive scheme: (1) Atomic Aesthetic Capability Learning, which strengthens AesRM's recognition of fundamental aesthetic concepts, e.g., accurately identifying centered composition; (2) Cold-Start, aligning the model with structured reasoning protocols; and (3) GRPO, further improving evaluation accuracy. To enhance AesRM-CoT, we additionally propose self-consistency-based CoT synthesis to improve CoT quality and design CoT-based process rewards during GRPO. Extensive experiments show AesRM outperforms baselines on multiple aesthetics benchmarks and is more robust, with lower position bias. Finally, we align Wan2.2 with AesRM and observe clear aesthetic gains over existing aesthetic reward models.

cs.CV↗

arXiv2026

NestPipe: Large-Scale Recommendation Training on 1,500+ Accelerators via Nested Pipelining

Modern recommendation models have increased to trillions of parameters. As cluster scales expand to O(1k), distributed training bottlenecks shift from computation and memory to data movement, especially lookup and communication latency associated with embeddings. Existing solutions either optimize only one bottleneck or improve throughput by sacrificing training consistency. This paper presents NestPipe, a large-scale decentralized embedding training framework that tackles both bottlenecks while preserving synchronous training semantics. NestPipe exploits two hierarchical sparse parallelism opportunities through nested pipelining. At the inter-batch level, Dual-Buffer Pipelining (DBP) constructs a staleness-free five-stage pipeline through dual-buffer synchronization, mitigating lookup bottlenecks without embedding staleness. At the intra-batch level, we identify the embedding freezing phenomenon, which inspires Frozen-Window Pipelining (FWP) to overlap All2All communication with dense computation via coordinated stream scheduling and key-centric sample clustering. Experiments on production GPU and NPU clusters with 1,536 workers demonstrate that NestPipe achieves up to 3.06x speedup and 94.07% scaling efficiency.

cs.DC↗

arXiv2026

ClickAIXR: On-Device Multimodal Vision-Language Interaction with Real-World Objects in Extended Reality

We present ClickAIXR, a novel on-device framework for multimodal vision-language interaction with objects in extended reality (XR). Unlike prior systems that rely on cloud-based AI (e.g., ChatGPT) or gaze-based selection (e.g., GazePointAR), ClickAIXR integrates an on-device vision-language model (VLM) with a controller-based object selection paradigm, enabling users to precisely click on real-world objects in XR. Once selected, the object image is processed locally by the VLM to answer natural language questions through both text and speech. This object-centered interaction reduces ambiguity inherent in gaze- or voice-only interfaces and improves transparency by performing all inference on-device, addressing concerns around privacy and latency. We implemented ClickAIXR in the Magic Leap SDK (C API) with ONNX-based local VLM inference. We conducted a user study comparing ClickAIXR with Gemini 2.5 Flash and ChatGPT 5, evaluating usability, trust, and user satisfaction. Results show that latency is moderate and user experience is acceptable. Our findings demonstrate the potential of click-based object selection combined with on-device AI to advance trustworthy, privacy-preserving XR interactions. The source code and supplementary materials are available at: nanovis.org/ClickAIXR.html

cs.CV↗

arXiv2026

DeepEye: A Steerable Self-driving Data Agent System

Large Language Models (LLMs) have revolutionized natural language interaction with data. The "holy grail" of data analytics is to build autonomous Data Agents that can self-drive complex data analysis workflows. However, current implementations are still limited to linear "ChatBI" systems. These systems struggle with joint analysis across heterogeneous data sources (e.g., databases, documents, and data files) and often encounter "context explosion" in complex and iterative data analysis workflows. To address these challenges, we present DeepEye, a production-ready data agent system that adopts a workflow-centric architecture to ensure scalability and trustworthiness. DeepEye introduces a Unified Multimodal Orchestration protocol, enabling seamless integration of structured and unstructured data sources. To mitigate hallucinations, it employs Hierarchical Reasoning with context isolation, decomposing complex intents into autonomous AgentNodes and deterministic ToolNodes. Furthermore, DeepEye incorporates a database-inspired Workflow Engine (comprising a Compiler, Validator, Optimizer, and Executor) that guarantees structural correctness and accelerates execution via runtime topological optimization. In this demonstration, we showcase DeepEye's ability to orchestrate complex workflows to generate diverse multimodal outputs -- including Data Videos, Dashboards, and Analytical Reports -- highlighting its advantages in transparent execution, automated optimization, and human-in-the-loop reliability.

cs.DB↗

arXiv2026

XAttnRes: Cross-Stage Attention Residuals for Medical Image Segmentation

In the field of Large Language Models (LLMs), Attention Residuals have recently demonstrated that learned, selective aggregation over all preceding layer outputs can outperform fixed residual connections. We propose Cross-Stage Attention Residuals (XAttnRes), a mechanism that maintains a global feature history pool accumulating both encoder and decoder stage outputs. Through lightweight pseudo-query attention, each stage selectively aggregates from all preceding representations. To bridge the gap between the same-dimensional Transformer layers in LLMs and the multi-scale encoder-decoder stages in segmentation networks, XAttnRes introduces spatial alignment and channel projection steps that handle cross-resolution features with negligible overhead. When added to existing segmentation networks, XAttnRes consistently improves performance across four datasets and three imaging modalities. We further observe that XAttnRes alone, even without skip connections, achieves performance on par with the baseline, suggesting that learned aggregation can recover the inter-stage information flow traditionally provided by predetermined connections.

cs.CV↗

arXiv2026

SAG-Agent: Enabling Long-Horizon Reasoning in Strategy Games via Dynamic Knowledge Graphs

Most commodity software lacks accessible Application Programming Interfaces (APIs), requiring autonomous agents to interact solely through pixel-based Graphical User Interfaces (GUIs). In this API-free setting, large language model (LLM)-based agents face severe efficiency bottlenecks: limited to local visual experiences, they make myopic decisions and rely on inefficient trial-and-error, hindering both skill acquisition and long-horizon planning. To overcome these limitations, we propose SAG-Agent, an experience-driven learning framework that structures an agent's raw pixel-level interactions into a persistent State-Action Graph (SAG). SAG-Agent mitigates inefficient exploration by topologically linking functionally similar but visually distinct GUI states, constructing a rich neighborhood of experience that enables the agent to generalize from a diverse set of historical strategies. To facilitate long-horizon reasoning, we design a novel hybrid intrinsic reward mechanism based on the graph topology, combining a state-value reward for exploiting known high-value pathways with a novelty reward that encourages targeted exploration. This approach decouples strategic planning from pure discovery, allowing the agent to effectively value setup actions with delayed gratification. We evaluate SAG-Agent in two complex, open-ended GUI-based decision-making environments (Civilization V and Slay the Spire), demonstrating significant improvements in exploration efficiency and strategic depth over the state-of-the-art methods.

cs.AI↗