arXiv subjects

Yijia Xu

Publications and source records attributed to Yijia Xu.

At least 19 recordsLinked to original sources

arXiv2026

A framework for low-overhead quantum fault tolerance via spacetime lifting

Fault-tolerant quantum computation is inherently a spacetime problem, requiring not merely good static quantum error-correcting codes but also low-overhead protocols for protecting and manipulating encoded quantum information over time. Fault complexes provide a homological formalism for treating such protocols as single spacetime objects. Here we initiate the study of low-overhead fault complexes by introducing \emph{spacetime lifting}, a method that constructs fault complexes from symmetry-reduced product structures beyond standard foliation. We show that spacetime lifting yields fault complexes and in particular memory experiments with almost-linear fault distance in the total spacetime cost, which substantially outperforms existing constructions. We further endow fault complexes with the operational interpretation of measurement-based cluster state protocols and identify general conditions under which they realize fault-tolerant logical teleportation, showing that spacetime-lifted constructions combine favorable parameter scaling with operational realizations. Our work establishes a systematic route towards more efficient quantum fault tolerance through general complex constructions.

quant-ph↗

arXiv2026

How Effective Are NPM Malicious Package Detectors? A Large-Scale Empirical Study

The NPM ecosystem faces escalating threats from malicious packages that exploit its open publication model. While numerous detection tools have been proposed, they are evaluated on disparate datasets with inconsistent settings, making cross-tool comparison unreliable and leaving practitioners without clear guidance. We present the first large-scale empirical study of NPM malicious package detection, evaluating 11 tools with 16 variants on a unified benchmark of 6,420 malicious and 7,288 benign packages annotated with 11 behavior categories and 8 evasion techniques. Unlike prior work, we inspect each tool's source code to explain why tools succeed or fail, not merely how often. Our key findings: (1) the precision and recall a tool achieves are structurally determined by how it resolves the ambiguity between code capability and malicious intent, with IntelGuard reaching the best F1 at 95.98% by grounding its judgment in retrieved evidence and GuardDog the best among conventional tools at 93.32%; (2) behavioral coupling amplifies detection signals when behaviors co-occur, raising SAP_DT from 3.2% to 79.3% for the collect-and-exfiltrate chain; (3) 80.3% of malware uses no evasion because the ecosystem lacks mandatory pre-publication scanning; (4) ML degradation is driven by concept convergence rather than concept drift, since malware became simpler and every decision boundary fitted to a corpus ages with it; (5) combination effectiveness equals complementarity minus false-positive introduction, not paradigm diversity. Strategic combinations reach up to 97.21% accuracy and 97.02% F1. We release our benchmark and evaluation framework.

cs.SE↗

arXiv2026

LCG: Long-Context Consistent Image Generation with Sparse Relational Attention

Recent image generation models achieve impressive quality in single-image synthesis, but often fail to maintain consistency across sequential outputs, as required in comics, storyboards, and visual narratives. We propose Long-Context Generation (LCG), a framework for long-context multi-image text-to-image generation, to improve consistency and scalability in long-context multi-image generation. LCG employs the Sparse Relational Attention (SRA) mechanism to selectively attend to core features across extended visual contexts, ensuring that the propagation of semantic and layout information remains computationally tractable. To enforce semantic alignment, we introduce the Routing Consistency Constraint (RCC), which leverages identity-aware masks to align structural patterns across generation branches, effectively mitigating drift in appearance even in complex multi-character scenes. To support training and evaluation in this setting, we construct the Long-Context Consistency Dataset (LCCD), a large-scale synthetic dataset comprising character-centric multi-image sequences spanning varied situational contexts. LCCD contains 600K training sequences and a separate 1K test set, with each sequence containing 6 to 20 images. The experiments demonstrate that LCG outperforms the compared baselines in prompt alignment and character consistency for long-context image generation, including multi-character scenes.

cs.CV↗

arXiv2026

Hierarchical Concept-to-Appearance Guidance for Multi-Subject Image Generation

Multi-subject image generation aims to synthesize images that faithfully preserve the identities of multiple reference subjects while following textual instructions. However, existing methods often suffer from identity inconsistency and limited compositional control, as they rely on diffusion models to implicitly associate text prompts with reference images. In this work, we propose Hierarchical Concept-to-Appearance Guidance (CAG), a framework that provides explicit, structured supervision from high-level concepts to fine-grained appearances. At the conceptual level, we introduce a VAE dropout training strategy that randomly omits reference VAE features, encouraging the model to rely more on robust semantic signals from a Visual Language Model (VLM) and thereby promoting consistent concept-level generation in the absence of complete appearance cues. At the appearance level, we integrate the VLM-derived correspondences into a correspondence-aware masked attention module within the Diffusion Transformer (DiT). This module restricts each text token to attend only to its matched reference regions, ensuring precise attribute binding and reliable multi-subject composition. Extensive experiments demonstrate that our method achieves state-of-the-art performance on the multi-subject image generation, substantially improving prompt following and subject consistency.

cs.CV↗

arXiv2026

MalSkillBench: A Runtime-Verified Benchmark of Malicious Agent Skills

AI coding agents such as Claude Code and Gemini CLI increasingly extend themselves with third-party skills: markdown packages bundling natural-language instructions, executable scripts, and tool permissions. Because a skill is at once code and agent-facing instruction, it introduces a supply chain dependency whose risk is neither pure code nor pure prompt. Detection tools have never been measured against verified ground truth spanning this hybrid space, leaving their effectiveness unknown and wild-only evaluations biased. We present MalSkillBench, the first runtime-verified benchmark of malicious agent skills: 3,944 malicious skills labeled along a three-dimensional taxonomy of 108 cells. Of these, 3,214 come from a closed-loop Generate-Verify-Feedback pipeline admitting only samples whose malicious behavior fires inside a Docker sandbox under system-call monitoring and an LLM judge; we add 703 in-the-wild and 4,000 matched benign skills. Our measurements are consistent: code injection reaches 94.5% verification yield but prompt injection only 75.8%, the same fragility that later makes it hard to detect; the wild sample is narrow, dominated by one cryptocurrency-theft campaign (86.6% one behavior, 81% from two accounts) with a small but architecturally new tail attacking the agent control plane; the strongest skill-specific detector reaches 98.4% recall on code injection yet collapses on prompt-injection and agent-control attacks, and wild-only scoring swings the ranking by up to 66 recall points; supply-chain scanners and prompt-injection defenses each see only half of a skill, and no combination recovers the code-instruction relationship. Detecting malicious skills therefore requires reasoning jointly over task intent, code, and instructions. We release the dataset, pipeline, baselines, and results.

cs.CR↗

arXiv2026

Approximate quantum error correction theory of non-isometric codes

Non-isometric encoding arises in various important contexts in quantum error correction, most notably in the finite-energy, non-ideal codewords inevitable in experimental realizations of continuous-variable codes, and holographic quantum gravity. In this work, we present a general and systematic theory of non-isometric quantum error-correcting codes. In particular, we employ the approximate quantum error correction framework to quantitatively study the fundamental limitations imposed by non-isometric encodings on the accuracy of quantum error correction and implementation of logical operations. We apply our theory to analyze GKP and tiger codes under energy constraints, and discuss the implications to holography.

quant-ph↗

arXiv2026

Bosonic Cyclic Codes: Trading Stabilizers for Gaussian Non-Clifford Phase Gates

Bosonic codes offer hardware-efficient approaches to quantum error correction, with the best encodings offering effective protection of idle quantum information against loss and dephasing - particularly rotation-symmetric codes, which include the cat and binomial code families. However, rotation-symmetric codes are only naturally endowed with a single logical Pauli gate, while other logical gates require the use of non-linear operations, obstructing the utility of these codes for realizing quantum algorithms. Here, we balance error protection with controllability by introducing bosonic cyclic codes: a generalization of rotation-symmetric codes that enable the measured tradeoff of error protection properties for fault-tolerant logical phase gates. Through our general construction, we find that sacrificing the detectability of a single photon loss relative to a rotation-symmetric code can yield a number of logical phase gates commensurate with the original rotation symmetry order of the code, all achievable via passive Gaussian rotations. Giving the corresponding generalizations of cat and binomial codes - which we dub cyclic cat and Vandermonde codes, respectively - we further find that many of the desirable properties of these codes transfer to the bosonic cyclic code setting. We go on to discuss the larger $SU(2)$ symmetry and rotation gates of the codes, which yield additional stabilizers and logical Pauli gates, as well as new non-Clifford gates for the smallest `kitten' binomial code, and provide a new error detection protocol. Finally, we introduce a general paradigm for converting higher-order stabilizers to logical gates, as in our generalization of rotation-symmetric codes, and apply it to several multimode bosonic codes.

quant-ph↗

arXiv2026

Understanding Older Adults' Experiences of Support, Concerns, and Risks from Kinship-Role AI-Generated Influencers

AI-generated influencers are rapidly gaining popularity on Chinese short-video platforms, often adopting kinship-based roles such as AI grandchildren to attract older adults. Although this trend has raised public concern, little is known about the design strategies behind these influencers, how older adults experience them, and the benefits and risks involved. In this study, we combined social media analysis with interviews to unpack the above questions. Our findings show that influencers use both visual and conversational cues to enact kinship roles, prompting audiences to engage in kinship-based role-play. Interviews further show that these cues arouse emotional resonance, help fulfill older adults' informational and emotional needs, while also raising concerns about emotional displacement and unequal emotional investment. We highlight the complex relationship between virtual avatars and real family ties, shaped by broader sociocultural norms, and discuss how AI might strengthen social support for older adults while mitigating risks within cultural contexts.

cs.HC↗

arXiv2026

An Efficient Additive Kolmogorov-Arnold Transformer for Point-Level Maize Localization in Unmanned Aerial Vehicle Imagery

High-resolution UAV photogrammetry has become a key technology for precision agriculture, enabling centimeter-level crop monitoring and point-level plant localization. However, point-level maize localization in UAV imagery remains challenging due to (1) extremely small object-to-pixel ratios, typically less than 0.1%, (2) prohibitive computational costs of quadratic attention on ultra-high-resolution images larger than 3000 x 4000 pixels, and (3) agricultural scene-specific complexities such as sparse object distribution and environmental variability that are poorly handled by general-purpose vision models. To address these challenges, we propose the Additive Kolmogorov-Arnold Transformer (AKT), which replaces conventional multilayer perceptrons with Pade Kolmogorov-Arnold Network (PKAN) modules to enhance functional expressivity for small-object feature extraction, and introduces PKAN Additive Attention (PAA) to model multiscale spatial dependencies with reduced computational complexity. In addition, we present the Point-based Maize Localization (PML) dataset, consisting of 1,928 high-resolution UAV images with approximately 501,000 point annotations collected under real field conditions. Extensive experiments show that AKT achieves an average F1-score of 62.8%, outperforming state-of-the-art methods by 4.2%, while reducing FLOPs by 12.6% and improving inference throughput by 20.7%. For downstream tasks, AKT attains a mean absolute error of 7.1 in stand counting and a root mean square error of 1.95-1.97 cm in interplant spacing estimation. These results demonstrate that integrating Kolmogorov-Arnold representation theory with efficient attention mechanisms offers an effective framework for high-resolution agricultural remote sensing.

cs.CV↗

arXiv2025

Letting the tiger out of its cage: bosonic coding without concatenation

Continuous-variable cat codes are encodings into a single photonic or phononic mode that offer a promising avenue for hardware-efficient fault-tolerant quantum computation. Protecting information in a cat code requires measuring the mode's occupation number modulo two, but this can be relaxed to a linear occupation-number constraint using the alternative two-mode pair-cat encoding. We construct multimode codes with similar linear constraints using any two integer matrices satisfying a CSS-like homological condition of a quantum rotor code. Just like the pair-cat code, syndrome extraction can be performed in tandem with stabilizing dissipation using current superconducting-circuit designs. The framework includes codes with various finite- or infinite-dimensional codespaces, and codes with finite or infinite Fock-state support. It encompasses two-component cat, pair-cat, dual-rail, two-mode binomial, various bosonic repetition codes, and aspects of chi-squared encodings while also yielding codes from homological products, lattices, generalized coherent states, and algebraic varieties. Among our examples are analogues of repetition codes, the Shor code, and a surface-like code that is not a concatenation of a known cat code with the qubit surface code. Codewords are coherent states projected into a Fock-state subspace defined by an integer matrix, and their overlaps are governed by Gelfand-Kapranov-Zelevinsky hypergeometric functions.

quant-ph↗

arXiv2025

Tessellation codes: encoded quantum gates by geometric rotation

We utilize the symmetry groups of regular tessellations on two-dimensional surfaces of different constant curvatures, including spheres, Euclidean planes and hyperbolic planes, to encode a qubit or qudit into the physical degrees of freedom on these surfaces, which we call tessellation codes. We show that tessellation codes exhibit decent error correction properties by analysis via geometric considerations and the representation theory of the isometry groups on the corresponding surfaces. Interestingly, we demonstrate how this formalism enables the implementation of certain logical operations through geometric rotations of surfaces in real space, opening a new approach to logical quantum computation. We provide a variety of concrete constructions of such codes associated with different tessellations, which give rise to different logical groups. This formalism sheds a new light on quantum code and logical operation construction.

quant-ph↗

arXiv2025

Knowledge-guided machine learning for county-level corn yield prediction under drought

Remote sensing (RS) technique, enabling the non-contact acquisition of extensive ground observations, is a valuable tool for crop yield predictions. Traditional process-based models struggle to incorporate large volumes of RS data, and most users lack understanding of crop growth mechanisms. In contrast, machine learning (ML) models are often criticized as "black boxes" due to their limited interpretability. To address these limitations, we utilized Knowledge-Guided Machine Learning (KGML), a framework that leverages the strengths of both process-based and ML models. Existing works have either overlooked the role of soil moisture in corn growth or did not embed this effect into their models. To bridge this gap, we developed the Knowledge-Guided Machine Learning with Soil Moisture (KGML-SM) framework, treating soil moisture as an intermediate variable in corn growth to emphasize its key role in plant development. Additionally, based on the prior knowledge that the model may overestimate under drought conditions, we designed a drought-aware loss function that penalized predicted yield in drought-affected areas. Our experiments showed that the KGML-SM model outperformed other traditional ML models. We explored the relationships between drought, soil moisture, and corn yield prediction by assessing the importance of different features within the model, and analyzing how soil moisture impacts predictions across different regions and time periods. Finally we provided interpretability for prediction errors to guide future model optimization.

cs.LG↗

arXiv2025

The Agent Behavior: Model, Governance and Challenges in the AI Digital Age

Advancements in AI have led to agents in networked environments increasingly mirroring human behavior, thereby blurring the boundary between artificial and human actors in specific contexts. This shift brings about significant challenges in trust, responsibility, ethics, security and etc. The difficulty in supervising of agent behaviors may lead to issues such as data contamination and unclear accountability. To address these challenges, this paper proposes the "Network Behavior Lifecycle" model, which divides network behavior into 6 stages and systematically analyzes the behavioral differences between humans and agents at each stage. Based on these insights, the paper further introduces the "Agent for Agent (A4A)" paradigm and the "Human-Agent Behavioral Disparity (HABD)" model, which examine the fundamental distinctions between human and agent behaviors across 5 dimensions: decision mechanism, execution efficiency, intention-behavior consistency, behavioral inertia, and irrational patterns. The effectiveness of the model is verified through real-world cases such as red team penetration and blue team defense. Finally, the paper discusses future research directions in dynamic cognitive governance architecture, behavioral disparity quantification, and meta-governance protocol stacks, aiming to provide a theoretical foundation and technical roadmap for secure and trustworthy human-agent collaboration.

cs.AI↗

arXiv2025

A Vision for Auto Research with LLM Agents

This paper introduces Agent-Based Auto Research, a structured multi-agent framework designed to automate, coordinate, and optimize the full lifecycle of scientific research. Leveraging the capabilities of large language models (LLMs) and modular agent collaboration, the system spans all major research phases, including literature review, ideation, methodology planning, experimentation, paper writing, peer review response, and dissemination. By addressing issues such as fragmented workflows, uneven methodological expertise, and cognitive overload, the framework offers a systematic and scalable approach to scientific inquiry. Preliminary explorations demonstrate the feasibility and potential of Auto Research as a promising paradigm for self-improving, AI-driven research processes.

cs.AI↗

arXiv2025

Extracting topological orders of generalized Pauli stabilizer codes in two dimensions

In this paper, we introduce an algorithm for extracting topological data from translation invariant generalized Pauli stabilizer codes in two-dimensional systems, focusing on the analysis of anyon excitations and string operators. The algorithm applies to $\mathbb{Z}_d$ qudits, including instances where $d$ is a nonprime number. This capability allows the identification of topological orders that differ from the $\mathbb{Z}_d$ toric codes. It extends our understanding beyond the established theorem that Pauli stabilizer codes for $\mathbb{Z}_p$ qudits (with $p$ being a prime) are equivalent to finite copies of $\mathbb{Z}_p$ toric codes and trivial stabilizers. The algorithm is designed to determine all anyons and their string operators, enabling the computation of their fusion rules, topological spins, and braiding statistics. The method converts the identification of topological orders into computational tasks, including Gaussian elimination, the Hermite normal form, and the Smith normal form of truncated Laurent polynomials. Furthermore, the algorithm provides a systematic approach for studying quantum error-correcting codes. We apply it to various codes, such as self-dual CSS quantum codes modified from the 2d honeycomb color code and non-CSS quantum codes that contain the double semion topological order or the six-semion topological order.

quant-ph↗

arXiv2025

Direction-Aware Diagonal Autoregressive Image Generation

The raster-ordered image token sequence exhibits a significant Euclidean distance between index-adjacent tokens at line breaks, making it unsuitable for autoregressive generation. To address this issue, this paper proposes Direction-Aware Diagonal Autoregressive Image Generation (DAR) method, which generates image tokens following a diagonal scanning order. The proposed diagonal scanning order ensures that tokens with adjacent indices remain in close proximity while enabling causal attention to gather information from a broader range of directions. Additionally, two direction-aware modules: 4D-RoPE and direction embeddings are introduced, enhancing the model's capability to handle frequent changes in generation direction. To leverage the representational capacity of the image tokenizer, we use its codebook as the image token embeddings. We propose models of varying scales, ranging from 485M to 2.0B. On the 256$\times$256 ImageNet benchmark, our DAR-XL (2.0B) outperforms all previous autoregressive image generators, achieving a state-of-the-art FID score of 1.37.

cs.CV↗

arXiv2025

Clifford operations and homological codes for rotors and oscillators

We develop quantum information processing primitives for the planar rotor, the state space of a particle on a circle. By interpreting rotor wavefunctions as periodically identified wavefunctions of a harmonic oscillator, we determine the group of bosonic Gaussian operations inherited by the rotor. This $n$-rotor Clifford group, $\text{U}(1)^{n(n+1)/2} \rtimes \text{GL}_n(\mathbb{Z})$, is represented by continuous $\text{U}(1)$ gates generated by polynomials quadratic in angular momenta, as well as discrete $\text{GL}_n(\mathbb Z)$ momentum sign-flip and sum gates. We classify homological rotor error-correcting codes [arXiv:2303.13723] and various rotor states based on equivalence under Clifford operations. Reversing direction, we map homological rotor codes and rotor Clifford operations back into oscillators by interpreting occupation-number states as rotor states of non-negative angular momentum. This yields new multimode homological bosonic codes protecting against dephasing and changes in occupation number, along with their corresponding encoding and decoding circuits. In particular, we show how to non-destructively measure the oscillator phase using conditional occupation-number addition and post selection. We also outline several rotor and oscillator varieties of the GKP-stabilizer codes [arXiv:1903.12615].

quant-ph↗

arXiv2024

SpearBot: Leveraging Large Language Models in a Generative-Critique Framework for Spear-Phishing Email Generation

Large Language Models (LLMs) are increasingly capable, aiding in tasks such as content generation, yet they also pose risks, particularly in generating harmful spear-phishing emails. These emails, crafted to entice clicks on malicious URLs, threaten personal information security. This paper proposes an adversarial framework, SpearBot, which utilizes LLMs to generate spear-phishing emails with various phishing strategies. Through specifically crafted jailbreak prompts, SpearBot circumvents security policies and introduces other LLM instances as critics. When a phishing email is identified by the critic, SpearBot refines the generated email based on the critique feedback until it can no longer be recognized as phishing, thereby enhancing its deceptive quality. To evaluate the effectiveness of SpearBot, we implement various machine-based defenders and assess how well the phishing emails generated could deceive them. Results show these emails often evade detection to a large extent, underscoring their deceptive quality. Additionally, human evaluations of the emails' readability and deception are conducted through questionnaires, confirming their convincing nature and the significant potential harm of the generated phishing emails.

cs.CR↗