arXiv subjects

Zhihao Xu

Publications and source records attributed to Zhihao Xu.

At least 37 records · Page 2Linked to original sources

arXiv2026

SAVE: Speech-Aware Video Representation Learning for Video-Text Retrieval

For video-text retrieval, the use of CLIP has been a de facto choice. Since CLIP provides only image and text encoders, this consensus has led to a biased paradigm that entirely ignores the sound track of videos. While several attempts have been made to reintroduce audio -- typically by incorporating an audio encoder and fusing its output with visual features -- these methods face two challenges: ineffective representation of speech content and suboptimal vision-audio fusion. To address these issues jointly, we propose SAVE, a Speech Aware Video rEpresentation learning method. SAVE improves upon AVIGATE, a SOTA audiovisual method, with a dedicated speech branch for more effective speech embedding. Furthermore, we introduce soft-ALBEF for early vision-audio alignment that facilitates fusion. Extensive experiments on five benchmarks show that SAVE compares favorably against the SOTA, outperforming AVIGATE by +4.1% on MSRVTT-9k, +1.9% on MSRVTT-7k, +2.5% on VATEX, +9.8% on Charades, and +2.1% on LSMDC, in light of the SumR metric.

cs.CV↗

arXiv2026

AOrchestra: Automating Sub-Agent Creation for Agentic Orchestration

Language agents have shown strong promise for task automation. Realizing this promise for increasingly complex, long-horizon tasks has driven the rise of a sub-agent-as-tools paradigm for multi-turn task solving. However, existing designs still lack a dynamic abstraction view of sub-agents, thereby hurting adaptability. We address this challenge with a unified, framework-agnostic agent abstraction that models any agent as a tuple Instruction, Context, Tools, Model. This tuple acts as a compositional recipe for capabilities, enabling the system to spawn specialized executors for each task on demand. Building on this abstraction, we introduce an agentic system AOrchestra, where the central orchestrator concretizes the tuple at each step: it curates task-relevant context, selects tools and models, and delegates execution via on-the-fly automatic agent creation. Such designs enable reducing human engineering efforts, and remain framework-agnostic with plug-and-play support for diverse agents as task executors. It also enables a controllable performance-cost trade-off, allowing the system to approach Pareto-efficient. Across three challenging benchmarks (GAIA, SWE-Bench, Terminal-Bench), AOrchestra achieves 16.28% relative improvement against the strongest baseline when paired with Gemini-3-Flash. The code is available at: https://github.com/FoundationAgents/AOrchestra

cs.AI↗

arXiv2026

Interaction-induced moiré lattices: from mosaic mobility edges to many-body localization

We study localization driven solely by interparticle interactions in moiré lattice systems without intrinsic disorder or externally imposed quasiperiodic potentials. We consider a one-dimensional bilayer with incommensurate lattice constants, described by a spin-dependent Fermi-Hubbard-type model with short-range interlayer interactions, where quasiperiodicity emerges only through interactions. Exact diagonalization shows that quenching hopping in one layer generates an interaction-induced mosaic potential with multiple mobility edges. When both layers are dynamical, increasing interlayer interactions drives transitions among ergodic, critical, and many-body localized regimes, with energy-dependent coexistence in certain parameter ranges. An exact mapping to a noninteracting single-particle model on a higher-dimensional structured graph provides a unified interpretation of these results and suggests an experimentally accessible route to interaction-induced moiré physics and localization.

cond-mat.dis-nn↗

arXiv2026

Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text

Enabling Large Language Models (LLMs) to effectively utilize tools in multi-turn interactions is essential for building capable autonomous agents. However, acquiring diverse and realistic multi-turn tool-use data remains a significant challenge. In this work, we propose a novel text-based paradigm. We observe that textual corpora naturally contain rich, multi-step problem-solving experiences, which can serve as an untapped, scalable, and authentic data source for multi-turn tool-use tasks. Based on this insight, we introduce GEM, a data synthesis pipeline that enables the generation and extraction of multi-turn tool-use trajectories from text corpora through a four-stage process: relevance filtering, workflow & tool extraction, trajectory grounding, and complexity refinement. To reduce the computational cost, we further train a specialized Trajectory Synthesizer via supervised fine-tuning. This model distills the complex generation pipeline into an efficient, end-to-end trajectory generator. Experiments demonstrate that our GEM-32B achieve a 16.5% improvement on the BFCL V3 Multi-turn benchmark. Our models partially surpass the performance of models trained on τ - bench (Airline and Retail) in-domain data, highlighting the superior generalization capability derived from our text-based synthesis paradigm. Notably, our Trajectory Synthesizer matches the quality of the full pipeline while significantly reducing inference latency and costs.

cs.CL↗

arXiv2025

Database and deep-learning scalability of anharmonic phonon properties by automated brute-force first-principles calculations

Understanding the anharmonic phonon properties of crystal compounds -- such as phonon lifetimes and thermal conductivities -- is essential for investigating and optimizing their thermal transport behaviors. These properties also impact optical, electronic, and magnetic characteristics through interactions between phonons and other quasiparticles and fields. In this study, we develop an automated first-principles workflow to calculate anharmonic phonon properties and build a comprehensive database encompassing more than 6,000 inorganic compounds. Utilizing this dataset, we train a graph neural network model to predict thermal conductivity values and spectra from structural parameters, demonstrating a scaling law in which prediction accuracy improves with increasing training data size. High-throughput screening with the model enables the identification of materials exhibiting extreme thermal conductivities -- both high and low. The resulting database offers valuable insights into the anharmonic behavior of phonons, thereby accelerating the design and development of advanced functional materials.

cond-mat.mtrl-sci↗

arXiv2025

Size-Dependent Skin Effect Transitions in Weakly Coupled Non-Reciprocal Chains

Non-Hermitian systems exhibit unique boundary phenomena absent in their Hermitian counterparts, most notably the non-Hermitian skin effect (NHSE). In this work, we explore a lattice model consisting of two coupled non-reciprocal chains, focusing on the interplay between system size, inter-chain coupling, and spectral topology. Using both analytical and numerical approaches, we systematically examine the evolution of the complex energy spectra and spectral winding numbers under periodic and open boundary conditions. Our results uncover a variety of size-dependent localization transitions, including the emergence and instability of concurrent bipolar skin effects in the $W=0$ region, and their crossover to unipolar and conventional bipolar NHSE as the system size increases. Notably, we demonstrate that these size-dependent behaviors persist even beyond the weak-coupling regime, highlighting their universality in non-Hermitian systems with complex spectral structures. This study provides insights into the mechanisms governing skin effects and offers practical guidelines for engineering non-Hermitian topological phases in synthetic lattices.

physics.optics↗

arXiv2025

Topological Phase Transitions and Edge-State Transfer in Time-Multiplexed Quantum Walks

We investigate the topological phase transitions and edge-state properties of a time-multiplexed nonunitary quantum walk with sublattice symmetry. By constructing a Floquet operator incorporating tunable gain and loss, we systematically analyze both unitary and nonunitary regimes. In the unitary case, the conventional bulk-boundary correspondence (BBC) is preserved, with edge modes localized at opposite boundaries as predicted by topological invariants. In contrast, the nonunitary regime exhibits non-Hermitian skin effects, leading to a breakdown of the conventional BBC. By applying non-Bloch band theory and generalized Brillouin zones, we restore a generalized BBC and reveal a transfer phenomenon, where edge modes with different sublattice symmetries can become localized at the same boundary. Furthermore, we demonstrate that the structure of the spectral loops in the complex quasienergy plane provides a clear signature for these transfer behaviors. Our findings deepen the understanding of nonunitary topological phases and offer valuable insights for the experimental realization and control of edge states in non-Hermitian quantum systems.

quant-ph↗

arXiv2025

Quantitative correspondence between quasi-symmetric mappings on complete metric spaces and rough quasi-isometric mappings on their hyperbolic fillings

In this paper, we establish a quantitative correspondence between power quasi-symmetric mappings on complete metric spaces and rough quasi-isometric mappings on their hyperbolic fillings. In particular, we prove that the exponents in the power quasi-symmetric mappings coincide with the coefficients in the rough quasi-isometric mappings. This shows that the obtained correspondence is both sharp and consistent. In this way, we generalize the corresponding result by Björn, Björn, Gill, and Shanmugalingam (J. Reine Angew. Math., 2017) from the setting of rooted trees to that of hyperbolic fillings.

math.CV↗

arXiv2025

Reconstruction and Prediction of Volterra Integral Equations Driven by Gaussian Noise

Integral equations are widely used in fields such as applied modeling, medical imaging, and system identification, providing a powerful framework for solving deterministic problems. While parameter identification for differential equations has been extensively studied, the focus on integral equations, particularly stochastic Volterra integral equations, remains limited. This research addresses the parameter identification problem, also known as the equation reconstruction problem, in Volterra integral equations driven by Gaussian noise. We propose an improved deep neural networks framework for estimating unknown parameters in the drift term of these equations. The network represents the primary variables and their integrals, enhancing parameter estimation accuracy by incorporating inter-output relationships into the loss function. Additionally, the framework extends beyond parameter identification to predict the system's behavior outside the integration interval. Prediction accuracy is validated by comparing predicted and true trajectories using a 95% confidence interval. Numerical experiments demonstrate the effectiveness of the proposed deep neural networks framework in both parameter identification and prediction tasks, showing robust performance under varying noise levels and providing accurate solutions for modeling stochastic systems.

stat.ML↗

arXiv2025

Learning to Detect Unknown Jailbreak Attacks in Large Vision-Language Models

Despite extensive alignment efforts, Large Vision-Language Models (LVLMs) remain vulnerable to jailbreak attacks, posing serious safety risks. To address this, existing detection methods either learn attack-specific parameters, which hinders generalization to unseen attacks, or rely on heuristically sound principles, which limit accuracy and efficiency. To overcome these limitations, we propose Learning to Detect (LoD), a general framework that accurately detects unknown jailbreak attacks by shifting the focus from attack-specific learning to task-specific learning. This framework includes a Multi-modal Safety Concept Activation Vector module for safety-oriented representation learning and a Safety Pattern Auto-Encoder module for unsupervised attack classification. Extensive experiments show that our method achieves consistently higher detection AUROC on diverse unknown attacks while improving efficiency. The code is available at https://anonymous.4open.science/r/Learning-to-Detect-51CB.

cs.CV↗

arXiv2025

Goos-H$\ddot{a}$nchen shifts of bilayer meta-grating with unidirectional guide resonance

Bilayer meta-gratings with asymmetric structural parameters could host unidirectional guide resonances. The distribution of unidirectional guide resonances in the space of structural parameters and synthetic parameters is identified. As the incident optical beam being resonant with the unidirectional guide resonance, the Goos-H$\ddot{a}$nchen shifts of the scattered beams exhibit two anomalous behaviors: the resonant peak of the Goos-H$\ddot{a}$nchen shift is accompanied by constant transmittance and reflectance; the magnitude of the Goos-H$\ddot{a}$nchen shift is not always proportional to the quality factor of the unidirectional guide resonance. The temporal coupled mode theory analysis reveals that the first anomalous behavior is due to interference between direct scattering and radiation from the unidirectional guide resonance; the Goos-H$\ddot{a}$nchen shifts are proportional to the group velocity as well as the quality factor of the unidirectional guide resonance. Numerical simulations of incidence of Gaussian beam with finite beam width provide intuitive visualization of the Goos-H$\ddot{a}$nchen shift.

physics.optics↗

arXiv2025

OrthAlign: Orthogonal Subspace Decomposition for Non-Interfering Multi-Objective Alignment

Large language model (LLM) alignment faces a critical dilemma when addressing multiple human preferences: improvements in one dimension frequently come at the expense of others, creating unavoidable trade-offs between competing objectives like helpfulness and harmlessness. While prior work mainly focuses on constraint-based optimization algorithms and data selection strategies to mitigate conflicts, these approaches overlook the fundamental issue of resolving conflicts directly at the parameter level. In this paper, we present OrthAlign, an innovative approach that pioneers a new paradigm by leveraging orthogonal subspace decomposition to fundamentally resolve gradient-level conflicts in multi-objective preference alignment. OrthAlign strategically decomposes parameter update spaces into orthogonal subspaces, ensuring that optimization toward different preferences occurs in mathematically non-interfering directions. Building upon this, we provide theoretical guarantees demonstrating that when parameter increments satisfy both orthogonal subspace constraints and spectral norm bounds, the resulting updates exhibit linear Lipschitz growth rather than exponential instability, ensuring stable convergence across all preference dimensions. Extensive experiments show that: I. OrthAlign achieves maximum single-preference improvements ranging from 34.61% to 50.89% after multiple-objective alignment across helpful, harmless, and truthful dimensions. II. With an average overall reward improvement of 13.96%.

cs.LG↗

arXiv2025

Structural Mutation Based Differential Testing for FPGA Logic Synthesis Compilers

Field Programmable Gate Arrays (FPGAs) play a crucial role in Electronic Design Automation (EDA) applications, which have been widely used in safety-critical environments, including aerospace, chip manufacturing, and medical devices. A critical step in FPGA development is logic synthesis, which enables developers to translate their software designs into hardware net lists, which facilitates the physical implementation of the chip, detailed timing and power analysis, gate-level simulation, test vector generation, and optimization and consistency checking. However, bugs or incorrect implementations in FPGA logic synthesis compilers may lead to unexpected behaviors in target wapplications, posing security risks. Therefore, it is crucial to eliminate such bugs in FPGA logic synthesis compilers. The effectiveness of existing works is still limited by its simple, blind mutation strategy. To address this challenge, we propose a guided mutation strategy based on Bayesian optimization called LSC-Fuzz to detect bugs in FPGA logic synthesis compilers. Specifically, LSC-Fuzz consists of three components: the test-program generation component, the Bayesian diversity selection component, and the equivalent check component. By performing test-program generation and Bayesian diversity selection, LSC-Fuzz generates diverse and complex HDL code, thoroughly testing the FPGA logic synthesis compilers using equivalent check to detect bugs. Through three months, LSC-Fuzz has found 16 bugs, 12 of these has been confirmed by official technical support.

cs.SE↗

arXiv2025

Compiler Bugs Detection in Logic Synthesis Tools via Linear Upper Confidence Bound

Field-Programmable Gate Arrays (FPGAs) play an indispensable role in Electronic Design Automation (EDA), translating Register-Transfer Level (RTL) designs into gate-level netlists. The correctness and reliability of FPGA logic synthesis tools are critically important, as unnoticed bugs in these tools may infect the final hardware implementations. However, recent approaches often rely heavily on random selection strategies, limiting the structural diversity of the generated HDL test cases and resulting in inadequate exploration of the tool's feature space. To address this limitation, we propose Lin-Hunter, a novel testing framework designed to systematically enhance the diversity of HDL test cases and the efficiency of FPGA logic synthesis tool validation. Specifically, Lin-Hunter introduces a principled set of metamorphic transformation rules to generate functionally equivalent yet structurally diverse HDL test case variants, effectively addressing the limited diversity of existing test inputs. To further enhance bug discovery efficiency, Lin-Hunter integrates an adaptive strategy selection mechanism based on the Linear Upper Confidence Bound (LinUCB) method. This method leverages feedback from synthesis logs of previously executed test cases to dynamically prioritize transformation strategies that have empirically demonstrated a higher likelihood of triggering synthesis bugs. Comprehensive experiments conducted over a three-month period demonstrate the practical effectiveness of Lin-Hunter. Our method has discovered 18 unique bugs, including 10 previously unreported defects, which have been confirmed by official developers. Moreover, our method outperforms state-of-the-art testing methods in both test-case diversity and bug-discovery efficiency.

cs.SE↗

arXiv2025

Fibonacci-Modulation-Induced Multiple Topological Anderson Insulators

Topological Anderson insulators (TAIs) provide a mechanism for topological phase transitions in disordered systems and have implications for quantum material design. In this work, we investigate the emergence of multiple TAIs in a one-dimensional spin-orbit coupled (SOC) chain subject to Fibonacci modulation, which transforms a trivial band structure into a sequence of topologically nontrivial phases. This behavior is characterized by the appearance of zero-energy modes and changes in the $Z_2$ topological quantum number. As the SOC amplitude decreases, the number of TAI phases increases, a feature that is closely related to the fractal structure of the energy spectrum induced by Fibonacci modulation. In contrast to conventional TAI phases with fully localized eigenstates, the wave functions in the Fibonacci-modulated TAI phases display multifractal properties. This model can be experimentally realized using a Bose-Einstein condensate in a momentum-space lattice, where its topological transitions and multifractal features can be explored through quench dynamics.

cond-mat.dis-nn↗

arXiv2025

Rethinking LLM-Based RTL Code Optimization Via Timing Logic Metamorphosis

Register Transfer Level(RTL) code optimization is crucial for achieving high performance and low power consumption in digital circuit design. However, traditional optimization methods often rely on manual tuning and heuristics, which can be time-consuming and error-prone. Recent studies proposed to leverage Large Language Models(LLMs) to assist in RTL code optimization. LLMs can generate optimized code snippets based on natural language descriptions, potentially speeding up the optimization process. However, existing approaches have not thoroughly evaluated the effectiveness of LLM-Based code optimization methods for RTL code with complex timing logic. To address this gap, we conducted a comprehensive empirical investigation to assess the capability of LLM-Based RTL code optimization methods in handling RTL code with complex timing logic. In this study, we first propose a new benchmark for RTL optimization evaluation. It comprises four subsets, each corresponding to a specific area of RTL code optimization. Then we introduce a method based on metamorphosis to systematically evaluate the effectiveness of LLM-Based RTL code optimization methods.Our key insight is that the optimization effectiveness should remain consistent for semantically equivalent but more complex code. After intensive experiments, we revealed several key findings. (1) LLM-Based RTL optimization methods can effectively optimize logic operations and outperform existing compiler-based methods. (2) LLM-Based RTL optimization methods do not perform better than existing compiler-based methods on RTL code with complex timing logic, particularly in timing control flow optimization and clock domain optimization. This is primarily attributed to the challenges LLMs face in understanding timing logic in RTL code. Based on these findings, we provide insights for further research in leveraging LLMs for RTL code optimization.

cs.SE↗

arXiv2025

Paper Summary Attack: Jailbreaking LLMs through LLM Safety Papers

The safety of large language models (LLMs) has garnered significant research attention. In this paper, we argue that previous empirical studies demonstrate LLMs exhibit a propensity to trust information from authoritative sources, such as academic papers, implying new possible vulnerabilities. To verify this possibility, a preliminary analysis is designed to illustrate our two findings. Based on this insight, a novel jailbreaking method, Paper Summary Attack (\llmname{PSA}), is proposed. It systematically synthesizes content from either attack-focused or defense-focused LLM safety paper to construct an adversarial prompt template, while strategically infilling harmful query as adversarial payloads within predefined subsections. Extensive experiments show significant vulnerabilities not only in base LLMs, but also in state-of-the-art reasoning model like Deepseek-R1. PSA achieves a 97\% attack success rate (ASR) on well-aligned models like Claude3.5-Sonnet and an even higher 98\% ASR on Deepseek-R1. More intriguingly, our work has further revealed diametrically opposed vulnerability bias across different base models, and even between different versions of the same model, when exposed to either attack-focused or defense-focused papers. This phenomenon potentially indicates future research clues for both adversarial methodologies and safety alignment.Code is available at https://github.com/233liang/Paper-Summary-Attack

cs.CL↗

arXiv2025

Internal Value Alignment in Large Language Models through Controlled Value Vector Activation

Aligning Large Language Models (LLMs) with human values has attracted increasing attention since it provides clarity, transparency, and the ability to adapt to evolving scenarios. In this paper, we introduce a Controlled Value Vector Activation (ConVA) method that directly aligns the internal values of LLMs by interpreting how a value is encoded in their latent representations and modifies relevant activations to ensure consistent values in LLMs. To ensure an accurate and unbiased interpretation, we propose a context-controlled value vector identification method. To consistently control values without sacrificing model performance, we introduce a gated value vector activation method for effective and minimum degree of value control. Experiments show that our method achieves the highest control success rate across 10 basic values without hurting LLM performance and fluency, and ensures target values even with opposite and potentially malicious input prompts. Source code and data are available at~ https://github.com/hr-jin/ConVA.

cs.CL↗