arXiv subjects

Kaiwen Xue

Publications and source records attributed to Kaiwen Xue.

At least 19 recordsLinked to original sources

arXiv2026

PreGress: Ranking-Native Pre-training and Prompting for Graph Node Ranking

Node ranking is a fundamental problem in graph information retrieval, measuring the relative importance of nodes and supporting a wide range of applications such as influence analysis, recommendation, and graph-based retrieval augmented generation. However, exact computation of graph-based ranking measures is often computationally prohibitive at scale. Existing GNN-based ranking methods provide scalable approximations, but they are typically tailored to individual ranking criteria and require retraining for each downstream task, which limits their transferability and efficiency. Recent graph pre-training approaches aim to enable knowledge transfer across tasks, yet their learning objectives are largely misaligned with node ranking, resulting in suboptimal adaptability to ranking-oriented applications. To address these limitations, we propose PreGress, the first ranking-native pre-training and prompting framework for supporting a wide range of node ranking tasks. PreGress performs multi-task pre-training using our carefully designed objectives, including degree centrality prediction and attribute reconstruction, to jointly capture structural and attribute information. To support heterogeneous ranking criteria, we design lightweight, task-specific prompt modules that adapt a frozen ranking backbone to downstream tasks without full retraining. Experiments on six public graphs and two real-world query-to-item benchmarks---Yelp2018 and MovieLens-100K---together with a controlled five-criterion graph-access study demonstrate strong ranking quality with low task-specific state overhead.

cs.IR↗

arXiv2026

Efficient Graph Indexing for Interval-Aware Vector Search

Interval-aware Approximate Nearest Neighbor (ANN) search arises in applications where each object is associated with a numeric value or interval, and queries must satisfy both vector-similarity and interval constraints. Existing methods are typically tailored to a single query semantics, such as interval-filtered ANN search, and therefore require multiple specialized indexes to support diverse workloads, leading to substantial indexing and memory overhead. To address this limitation, we propose the Unified Interval-aware Relative Neighborhood Graph (URNG), a unified graph framework for interval-aware ANN search. URNG preserves the monotonic searchability of relative-neighborhood-graph based ANN indexes while additionally ensuring structural heredity over query-induced subgraphs, enabling a single index to support multiple interval-aware query semantics. Building on this framework, we develop UG, a practical graph index that efficiently approximates URNG through unified interval-aware pruning and iterative repair, together with a query algorithm for interval-aware ANN search. Extensive experiments on 5 datasets show that UG consistently achieves a strong accuracy-efficiency trade-off across diverse interval-aware workloads while maintaining competitive index construction cost and memory usage.

cs.DB↗

arXiv2026

ERGeoBench:A Comprehensive Benchmark for Embodied Reasoning and Geo-localization in Multimodal Large Language Models

Multimodal large language models (MLLMs) have shown strong potential as embodied agents, yet embodied geo-localization remains underexplored due to the lack of fine-grained evaluation. We introduce ERGeoBench, a diagnostic benchmark for vision-driven embodied geo-localization. ERGeoBench evaluates models under three progressive settings -- single-view, panorama-view, and embodied-view -- where agents may actively acquire observations through sequential changes in yaw, pitch, and zoom. The benchmark contains 2,207 globally distributed street-view panoramas and measures four complementary capabilities: foundational perception, spatial awareness, common sense reasoning, and geo-localization reasoning. Evaluations of leading proprietary and open-source MLLMs show that current models can infer high-level geographic semantics, but still struggle with fine-grained perceptual operations, metric localization, and spatial consistency across views. We further observe that geo-localization is strongly correlated with the other capability dimensions, suggesting that accurate localization depends on integrated perception, spatial reasoning, and commonsense inference rather than isolated visual recognition. Overall, ERGeoBench provides a unified framework for diagnosing and advancing human-like embodied geo-localization. Project Page: https://kaixuewen.github.io/ERGeoBench/

cs.CV↗

arXiv2026

MCI: A Maximal Clique Index for Efficient Arbitrary-Filtered Approximate Nearest Neighbor Search

Approximate Nearest Neighbor Search with arbitrary filtering predicates (AFANNS) is essential for modern data applications, yet existing methods often incur substantial storage and computational costs. In this work, we introduce the Maximal Clique Index (\mci), a novel graph-based index designed for robust and efficient AFANNS. The core idea of \mci is to approximate a dense Nearest Neighbor Graph (NNG) through a compact, clique-based representation. We propose two key techniques: (1) Maximal Clique Cover (\mcc), which exploits the geometric transitivity of high-dimensional spaces to encode dense neighborhoods as maximal cliques, achieving an index with high compression and connectivity; and (2) Local Neighborhood Graph Geometric Densification, a strategy that constructs an index approximating a large NNG from a sparse initial NNG, recovers global connectivity by progressively increasing distance thresholds to locally densify the structure. The index is built in a lock-free parallel manner for scalability and queried via a carefully-designed multi-seed strategy to handle fragmented predicate-induced subgraphs. Extensive experiments on 10 datasets show that \mci significantly outperforms state-of-the-art methods by up to one order of magnitude in QPS at high recall while using substantially smaller space, and remains competitive even on range/keyword filtering tasks, demonstrating robust general-purpose performance.

cs.DB↗

arXiv2026

Homodyne Photonic Tensor Processor exceeds 1,000-TOPS

High-performance computing underpins modern artificial intelligence (AI), enabling foundation models, real-time inference and perception in autonomous systems, and data-intensive scientific simulations. Recent advances in quantization techniques utilizing low-precision computation without degrading model accuracy, create new opportunities for analog photonic computing characterized by ultra-high clock rates and low energy consumption. Here we propose and demonstrate a coherent homodyne integrated circuit capable of general matrix multiplication (GEMM) with aggregate throughput that exceeds 1,000 TOPS (tera-operations per second), enabled by massive on-chip optical fanout and parallelism. By leveraging time multiplexing, the required modulator count is reduced from O($N^2$) to O(N), allowing dense integration of record-scale 256 $\times$ 256 homodyne units (each <0.0064 $mm^2$) within a single reticle. We employ wafer-scale fabricated 64 thin-film lithium niobate (TFLN) transmitters (each over 40-GHz bandwidth with propagation loss of 0.2 dB/cm) to encode data and chip-to-chip coupled to Si/SiN computing circuits (64 channels). Our system achieves up to 7-bit computational accuracy across 8 $\times$ 8 parallel channels at record computing clockrate 120 Gbaud/s, and 6-bit statistical accuracy across 256 $\times$ 100 channels at 20-128 Gbaud/s, representing a total throughput of 1,000-6,000 TOPS. Massive parallelism amortizes the optoelectronic (OE) conversion to allow 330-TOPS/W efficiency using foundry-available packaging technology. The system throughput is benchmarked with Qwen2.5-0.5 billion parameter models that generate accurate tokens. High throughput and energy efficiency establish a near-term pathway toward light-based accelerators for large-scale training and low-latency inference from datacenters to edges, accelerating new models toward artificial general intelligence.

cs.ET↗

arXiv2026

Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Data

Discrete diffusion models with absorbing processes have shown promise in language modeling. The key quantities to be estimated are the ratios between the marginal probabilities of two transitive states at all timesteps, called the concrete score. In this paper, we reveal that the concrete score in absorbing diffusion can be expressed as conditional probabilities of clean data, multiplied by a time-dependent scalar in an analytic form. Motivated by this finding, we propose reparameterized absorbing discrete diffusion (RADD), a dedicated diffusion model without time-condition that characterizes the time-independent conditional probabilities. Besides its simplicity, RADD can reduce the number of function evaluations (NFEs) by caching the output of the time-independent network when the noisy sample remains unchanged in a sampling interval, which enables sampling acceleration. Built upon the new perspective of conditional distributions, we further unify absorbing discrete diffusion and any-order autoregressive models (AO-ARMs), showing that the upper bound on the negative log-likelihood for the diffusion model can be interpreted as an expected negative log-likelihood for AO-ARMs. Further, our RADD models achieve SOTA performance among diffusion models on 5 zero-shot language modeling benchmarks (measured by perplexity) at the GPT-2 scale. Our code is available at https://github.com/ML-GSAI/RADD.

cs.LG↗

arXiv2026

Preserving Full Degradation Details for Blind Image Super-Resolution

The performance of image super-resolution relies heavily on the accuracy of degradation information, especially under blind settings. Due to the absence of true degradation models in real-world scenarios, previous methods learn distinct representations by distinguishing different degradations in a batch. However, the most significant degradation differences may provide shortcuts for the learning of representations such that subtle difference may be discarded. In this paper, we propose an alternative to learn degradation representations through reproducing degraded low-resolution (LR) images. By guiding the degrader to reconstruct input LR images, full degradation information can be encoded into the representations. In addition, we develop a distribution alignment loss to facilitate the learning of the degradation representations by introducing a bounded constraint. Moreover, to achieve larger receptive fields to capture information from a wider region for better restoration results, we introduce a degradation-aware Mamba module to efficiently model long-range dependency between the anchor pixel and the surrounding informative pixels. And the module strikes a flexible adaption to various degradations based on the learned representations. Experiments show that our representations can extract accurate and highly robust degradation information. Evaluations on both synthetic and real images demonstrate that our ReDSR achieves state-of-the-art performance for the blind SR tasks.

cs.CV↗

arXiv2026

Quantization-aware Photonic Homodyne computing for Accelerated Artificial Intelligence and Scientific Simulation

Modern problems in high-performance computing, ranging from training and inferencing deep learning models in computer vision and language models to simulating complex physical systems with nonlinearly-coupled equations, require exponential growth of computational resources. Photonic analog systems are emerging with solutions of intrinsic parallelism, high bandwidth, and low propagation loss. However, their application has been hindered by the low analog accuracy due to the electro-optic distortion, material nonlinearities, and signal-to-noise ratios. Here we overcome this barrier with a quantization-aware digital-photonic mixed-precision framework across chiplets for accelerated AI processing and physical simulation. Using Lithium Niobate photonics with channel equalization techniques, we demonstrate linear multiplication (9-bit amplitude-phase decoupling) in homodyne optical logics with 6-bit precision at the clock rate of 128 giga-symbol-per-second (128 GS/s), enabling AI processing with 6 ns latency. Codesign hardware-algorithms, including iterative solvers, sparse-dense quantization, and bit-sliced matrix multiplication, explore photonic amplitude and phase coherence for complex-valued, physics-inspired computation. In electromagnetic problems, our approach yields 12-bit solutions for partial differential equations (PDEs) in scattering problems that would conventionally require up to 32-bit and often even 64-bit precision. These results preserve digital-level fidelity while leveraging the high-speed low-energy photonic hardware, establishing a pathway toward general-purpose optical acceleration for generative artificial intelligence, real-time robotics, and accurate simulation for climate challenges and biological discoveries.

cs.ET↗

arXiv2025

CreBench: Human-Aligned Creativity Evaluation from Idea to Process to Product

Human-defined creativity is highly abstract, posing a challenge for multimodal large language models (MLLMs) to comprehend and assess creativity that aligns with human judgments. The absence of an existing benchmark further exacerbates this dilemma. To this end, we propose CreBench, which consists of two key components: 1) an evaluation benchmark covering the multiple dimensions from creative idea to process to products; 2) CreMIT (Creativity Multimodal Instruction Tuning dataset), a multimodal creativity evaluation dataset, consisting of 2.2K diverse-sourced multimodal data, 79.2K human feedbacks and 4.7M multi-typed instructions. Specifically, to ensure MLLMs can handle diverse creativity-related queries, we prompt GPT to refine these human feedbacks to activate stronger creativity assessment capabilities. CreBench serves as a foundation for building MLLMs that understand human-aligned creativity. Based on the CreBench, we fine-tune open-source general MLLMs, resulting in CreExpert, a multimodal creativity evaluation expert model. Extensive experiments demonstrate that the proposed CreExpert models achieve significantly better alignment with human creativity evaluation compared to state-of-the-art MLLMs, including the most advanced GPT-4V and Gemini-Pro-Vision.

cs.AI↗

arXiv2025

High Clockrate Free-space Optical In-Memory Computing

The ability to process and act on data in real time is increasingly critical for applications ranging from autonomous vehicles, three-dimensional environmental sensing and remote robotics. However, the deployment of deep neural networks (DNNs) in edge devices is hindered by the lack of energy-efficient scalable computing hardware. Here, we introduce a fanout spatial time-of-flight optical neural network (FAST-ONN) that calculates billions of convolutions per second with ultralow latency and power consumption. This is enabled by the combination of high-speed dense arrays of vertical-cavity surface-emitting lasers (VCSELs) for input modulation with spatial light modulators of high pixel counts for in-memory weighting. In a three-dimensional optical system, parallel differential readout allows signed weight values accurate inference in a single shot. The performance is benchmarked with feature extraction in You-Only-Look-Once (YOLO) for convolution at 100 million frames per second (MFPS), and in-system backward propagation training with photonic reprogrammability. The VCSEL transmitters are implementable in any free-space optical computing systems to improve the clockrate to over gigahertz. The high scalability in device counts and channel parallelism enables a new avenue to scale up free space computing hardware.

cs.ET↗

arXiv2025

Enhancing Technical Documents Retrieval for RAG

In this paper, we introduce Technical-Embeddings, a novel framework designed to optimize semantic retrieval in technical documentation, with applications in both hardware and software development. Our approach addresses the challenges of understanding and retrieving complex technical content by leveraging the capabilities of Large Language Models (LLMs). First, we enhance user queries by generating expanded representations that better capture user intent and improve dataset diversity, thereby enriching the fine-tuning process for embedding models. Second, we apply summary extraction techniques to encode essential contextual information, refining the representation of technical documents. To further enhance retrieval performance, we fine-tune a bi-encoder BERT model using soft prompting, incorporating separate learning parameters for queries and document context to capture fine-grained semantic nuances. We evaluate our approach on two public datasets, RAG-EDA and Rust-Docs-QA, demonstrating that Technical-Embeddings significantly outperforms baseline models in both precision and recall. Our findings highlight the effectiveness of integrating query expansion and contextual summarization to enhance information access and comprehension in technical domains. This work advances the state of Retrieval-Augmented Generation (RAG) systems, offering new avenues for efficient and accurate technical document retrieval in engineering and product development workflows.

cs.IR↗

arXiv2025

A Survey on Latent Reasoning

Large Language Models (LLMs) have demonstrated impressive reasoning capabilities, especially when guided by explicit chain-of-thought (CoT) reasoning that verbalizes intermediate steps. While CoT improves both interpretability and accuracy, its dependence on natural language reasoning limits the model's expressive bandwidth. Latent reasoning tackles this bottleneck by performing multi-step inference entirely in the model's continuous hidden state, eliminating token-level supervision. To advance latent reasoning research, this survey provides a comprehensive overview of the emerging field of latent reasoning. We begin by examining the foundational role of neural network layers as the computational substrate for reasoning, highlighting how hierarchical representations support complex transformations. Next, we explore diverse latent reasoning methodologies, including activation-based recurrence, hidden state propagation, and fine-tuning strategies that compress or internalize explicit reasoning traces. Finally, we discuss advanced paradigms such as infinite-depth latent reasoning via masked diffusion models, which enable globally consistent and reversible reasoning processes. By unifying these perspectives, we aim to clarify the conceptual landscape of latent reasoning and chart future directions for research at the frontier of LLM cognition. An associated GitHub repository collecting the latest papers and repos is available at: https://github.com/multimodal-art-projection/LatentCoT-Horizon/.

cs.CL↗

arXiv2025

PM-MOE: Mixture of Experts on Private Model Parameters for Personalized Federated Learning

Federated learning (FL) has gained widespread attention for its privacy-preserving and collaborative learning capabilities. Due to significant statistical heterogeneity, traditional FL struggles to generalize a shared model across diverse data domains. Personalized federated learning addresses this issue by dividing the model into a globally shared part and a locally private part, with the local model correcting representation biases introduced by the global model. Nevertheless, locally converged parameters more accurately capture domain-specific knowledge, and current methods overlook the potential benefits of these parameters. To address these limitations, we propose PM-MoE architecture. This architecture integrates a mixture of personalized modules and an energy-based personalized modules denoising, enabling each client to select beneficial personalized parameters from other clients. We applied the PM-MoE architecture to nine recent model-split-based personalized federated learning algorithms, achieving performance improvements with minimal additional training. Extensive experiments on six widely adopted datasets and two heterogeneity settings validate the effectiveness of our approach. The source code is available at \url{https://github.com/dannis97500/PM-MOE}.

cs.LG↗

arXiv2024

Residual Attention Single-Head Vision Transformer Network for Rolling Bearing Fault Diagnosis in Noisy Environments

Rolling bearings play a crucial role in industrial machinery, directly influencing equipment performance, durability, and safety. However, harsh operating conditions, such as high speeds and temperatures, often lead to bearing malfunctions, resulting in downtime, economic losses, and safety hazards. This paper proposes the Residual Attention Single-Head Vision Transformer Network (RA-SHViT-Net) for fault diagnosis in rolling bearings. Vibration signals are transformed from the time to frequency domain using the Fast Fourier Transform (FFT) before being processed by RA-SHViT-Net. The model employs the Single-Head Vision Transformer (SHViT) to capture local and global features, balancing computational efficiency and predictive accuracy. To enhance feature extraction, the Adaptive Hybrid Attention Block (AHAB) integrates channel and spatial attention mechanisms. The network architecture includes Depthwise Convolution, Single-Head Self-Attention, Residual Feed-Forward Networks (Res-FFN), and AHAB modules, ensuring robust feature representation and mitigating gradient vanishing issues. Evaluation on the Case Western Reserve University and Paderborn University datasets demonstrates the RA-SHViT-Net's superior accuracy and robustness in complex, noisy environments. Ablation studies further validate the contributions of individual components, establishing RA-SHViT-Net as an effective tool for early fault detection and classification, promoting efficient maintenance strategies in industrial settings. Keywords: rolling bearings, fault diagnosis, Vision Transformer, attention mechanism, noisy environments, Fast Fourier Transform (FFT)

cs.CV↗

arXiv2024

Hypermultiplexed Integrated-Photonics-based Tensor Optical Processor

The escalating data volume and complexity resulting from the rapid expansion of artificial intelligence (AI), internet of things (IoT) and 5G/6G mobile networks is creating an urgent need for energy-efficient, scalable computing hardware. Here we demonstrate a hypermultiplexed integratedphotonics-based tensor optical processor (HITOP) that can perform trillions of operations per second (TOPS) at the energy efficiency of 40 TOPS/W. Space-time-wavelength three-dimensional (3D) optical parallelism enables O($N^{2}$) operations per clock-cycle using O($N$) modulator devices. The system is built with wafer-fabricated III/V micron-scale lasers and high-speed thin-film Lithium-Niobate electro-optics for encoding at 10s femtojoule/symbol. Lasing threshold incorporates analog inline rectifier (ReLu) nonlinearity for low-latency activation. The system scalability is verified with machine learning models of 405,000 parameters. A combination of high clockrates, energy-efficient processing and programmability unlocks the potential of light for large-scale AI accelerators in applications ranging from training of large AI models to real-time decision making in edge deployment.

cs.ET↗

arXiv2024

Photorefractive and pyroelectric photonic memory and long-term stability in thin-film lithium niobate microresonators

The stability of the integrated photonic circuits is of critical importance for many applications that require high frequency precision or robust operation over time, such as optomechanical sensing, frequency conversion, optical communication, and quantum optics. Photonic memory is useful for low-energy optical computing and interconnects. Thin film lithium niobate (TFLN), as an emerging photonic platform, exhibits complex material properties including pyroelectric (PE) and photorefractive (PR) effects which could lead to intra-device drift and excess noise under different environmental or operating conditions as well as be utilized for building photonic memory. However, the long-term stability and memory effect of its optical properties has not been explored. In this paper, we discovered a long-lived change of optical refractive index as a result of light excitation and temporal temperature variation using Z-cut TFLN microresonators and reveal a strong dependence of the instability with the crystal orientation of the thin film form. The recovery time are measured to be over 10 hours. Leveraging the photonic memory with a long relaxation time, we realize optical trimming of the cavity resonance frequencies. Our result offers insights towards understanding the fundamental noise properties and dynamic behavior of the integrated TFLN material and devices.

physics.optics↗

arXiv2024

Unifying Bayesian Flow Networks and Diffusion Models through Stochastic Differential Equations

Bayesian flow networks (BFNs) iteratively refine the parameters, instead of the samples in diffusion models (DMs), of distributions at various noise levels through Bayesian inference. Owing to its differentiable nature, BFNs are promising in modeling both continuous and discrete data, while simultaneously maintaining fast sampling capabilities. This paper aims to understand and enhance BFNs by connecting them with DMs through stochastic differential equations (SDEs). We identify the linear SDEs corresponding to the noise-addition processes in BFNs, demonstrate that BFN's regression losses are aligned with denoise score matching, and validate the sampler in BFN as a first-order solver for the respective reverse-time SDE. Based on these findings and existing recipes of fast sampling in DMs, we propose specialized solvers for BFNs that markedly surpass the original BFN sampler in terms of sample quality with a limited number of function evaluations (e.g., 10) on both image and text datasets. Notably, our best sampler achieves an increase in speed of 5~20 times for free. Our code is available at https://github.com/ML-GSAI/BFN-Solver.

cs.LG↗

arXiv2024

SCALE: Self-Correcting Visual Navigation for Mobile Robots via Anti-Novelty Estimation

Although visual navigation has been extensively studied using deep reinforcement learning, online learning for real-world robots remains a challenging task. Recent work directly learned from offline dataset to achieve broader generalization in the real-world tasks, which, however, faces the out-of-distribution (OOD) issue and potential robot localization failures in a given map for unseen observation. This significantly drops the success rates and even induces collision. In this paper, we present a self-correcting visual navigation method, SCALE, that can autonomously prevent the robot from the OOD situations without human intervention. Specifically, we develop an image-goal conditioned offline reinforcement learning method based on implicit Q-learning (IQL). When facing OOD observation, our novel localization recovery method generates the potential future trajectories by learning from the navigation affordance, and estimates the future novelty via random network distillation (RND). A tailored cost function searches for the candidates with the least novelty that can lead the robot to the familiar places. We collect offline data and conduct evaluation experiments in three real-world urban scenarios. Experiment results show that SCALE outperforms the previous state-of-the-art methods for open-world navigation with a unique capability of localization recovery, significantly reducing the need for human intervention. Code is available at https://github.com/KubeEdge4Robotics/ScaleNav.

cs.RO↗