arXiv subjects

Jianxun Li

Publications and source records attributed to Jianxun Li.

8 recordsLinked to original sources

arXiv2026

Few to Big: Prototype Expansion Network via Diffusion Learner for Point Cloud Few-shot Semantic Segmentation

Few-shot 3D point cloud semantic segmentation aims to segment novel categories using a minimal number of annotated support samples. However, prototypes derived from the limited non-structural point cloud support set are often misaligned and have a small capacity, hindering effective gen eralization to novel categories. This stems from two core issues: i) the prototype possess limited representational capacity fails to cover the full intra-class diversity of a novel category, and ii) the prototypes suffer from misalignment with the query space due to the inter-set inconsistency between support and query sets. To address these issues, our work focuses on leveraging the few support samples to construct a well-aligned big-capacity prototype. Motivated by the powerful generative capabilities of diffusion models, we re-purpose its pre-trained conditional encoder to provide rich feature components for prototype ex pansion. Subsequently, a push-pull force aligns this expanded prototype towards the query feature space. Under this setup, we introduce the Prototype Expansion Network (PENet), a framework that constructs aligned big-capacity prototypes from two complementary feature sources. Specifically, PENet employs a dual-stream learner architecture: it retains a conventional fully supervised Intrinsic Learner (IL) to distill representative features, while introducing a novel Diffusion Learner (DL) to provide rich generalizable features. The resulting dual prototypes are then processed by a Prototype Assimilation Module (PAM), which adopts a push-pull attention block to align the prototypes with the query space. Furthermore, a Prototype Calibration Mechanism (PCM) regularizes the final big-capacity prototype to prevent semantic drift. Extensive experiments on the S3DIS and ScanNet datasets demonstrate that PENet outperforms state-of-the-art methods across various few-shot settings.

cs.CV↗

arXiv2026

Parallelizing Tool Execution and LLM Generation for Low-Latency Agent Serving

LLM-powered agents execute tasks through a sequential loop of model generation and tool execution. Today's serving systems serialize this loop, leaving tool latency exposed on the task critical path. This paper presents PASTE, a tool-aware agent-serving system that predicts concrete future tool invocations from recurring agent patterns and executes them speculatively while the LLM is still generating. PASTE isolates speculative results until confirmed by the LLM and jointly schedules tool execution and returning LLM sessions to avoid shifting bottlenecks to the GPU. Across deep research, coding, and scientific-agent workloads, PASTE reduces average task completion time by 43.5% and lowers observed tool latency by 1.8x.

cs.DC↗

arXiv2026

Renormalization group analysis for bosonization coefficients in half-odd-integer Kitaev spin chains

Based on a renormalization group (RG) analysis, we study the bosonization formulas in spin-$S$ Kitaev-Gamma and Kitaev-Heisenberg-Gamma chains in the $(K<0,Γ>0,J>0)$ parameter region, where $S$ is a half-odd integer. We find that the effects associated with the breaking of emergent continuous symmetries in bosonization formulas scale as $1/S$ in the large-$S$ limit, which is in qualitative agreement with DMRG numerical results for Kitaev-Gamma chains. In Kitaev-Heisenberg-Gamma chains, symmetry analysis reveals ten independent bosonization coefficients, five of which are predicted by the RG analysis to have no dependence on the Heisenberg coupling up to linear order. Our work may offer valuable input for determining magnetic ordering tendencies in two-dimensional Kitaev spin models within a quasi-one-dimensional approach.

cond-mat.str-el↗

arXiv2026

MLOW: Interpretable Low-Rank Frequency Magnitude Decomposition of Multiple Effects for Time Series Forecasting

Separating multiple effects in time series is fundamental yet challenging for time-series forecasting (TSF). However, existing TSF models cannot effectively learn interpretable multi-effect decomposition by their smoothing-based temporal techniques. Here, a new interpretable frequency-based decomposition pipeline MLOW captures the insight: a time series can be represented as a magnitude spectrum multiplied by the corresponding phase-aware basis functions, and the magnitude spectrum distribution of a time series always exhibits observable patterns for different effects. MLOW learns a low-rank representation of the magnitude spectrum to capture dominant trending and seasonal effects. We explore low-rank methods, including PCA, NMF, and Semi-NMF, and find that none can simultaneously achieve interpretable, efficient and generalizable decomposition. Thus, we propose hyperplane-nonnegative matrix factorization (Hyperplane-NMF). Further, to address the frequency (spectral) leakage restricting high-quality low-rank decomposition, MLOW enables a flexible selection of input horizons and frequency levels via a mathematical mechanism. Visual analysis demonstrates that MLOW enables interpretable and hierarchical multiple-effect decomposition, robust to noises. It can also enable plug-and-play in existing TSF backbones with remarkable performance improvement but minimal architectural modifications.

cs.LG↗

arXiv2025

Fourier Basis Mapping: A Time-Frequency Learning Framework for Time Series Forecasting

The integration of Fourier transform and deep learning opens new avenues for time series forecasting. We reconsider the Fourier transform from a basis functions perspective. Specifically, the real and imaginary parts of the frequency components can be regarded as the coefficients of cosine and sine basis functions at tiered frequency levels, respectively. We find that existing Fourier-based methods face inconsistent starting cycles and inconsistent series length issues. They fail to interpret frequency components precisely and overlook temporal information. Accordingly, the novel Fourier Basis Mapping (FBM) method addresses these issues by integrating time-frequency features through Fourier basis expansion and mapping in the time-frequency space. Our approach extracts explicit frequency features while preserving temporal characteristics. FBM supports plug-and-play integration with various types of neural networks by only adjusting the first initial projection layer for better performance. First, we propose FBM-L, FBM-NL, and FBM-NP to enhance linear, MLP-based, and Transformer-based models, respectively, demonstrating the effectiveness of time-frequency features. Next, we propose a synergetic model architecture, termed FBM-S, which decomposes the seasonal, trend, and interaction effects into three separate blocks, each designed to model time-frequency features in a specialized manner. Finally, we introduce several techniques tailored for time-frequency features, including interaction masking, centralization, patching, rolling window projection, and multi-scale down-sampling. The results are validated on diverse real-world datasets for both long-term and short-term forecasting tasks with SOTA performance.

cs.LG↗

arXiv2025

ServerlessLoRA: Minimizing Latency and Cost in Serverless Inference for LoRA-Based LLMs

Serverless computing has grown rapidly for serving Large Language Model (LLM) inference due to its pay-as-you-go pricing, fine-grained GPU usage, and rapid scaling. However, our analysis reveals that current serverless can effectively serve general LLM but fail with Low-Rank Adaptation (LoRA) inference due to three key limitations: 1) massive parameter redundancy among functions where 99% of weights are unnecessarily duplicated, 2) costly artifact loading latency beyond LLM loading, and 3) magnified resource contention when serving multiple LoRA LLMs. These inefficiencies lead to massive GPU wastage, increased Time-To-First-Token (TTFT), and high monetary costs. We propose ServerlessLoRA, a novel serverless inference system designed for faster and cheaper LoRA LLM serving. ServerlessLoRA enables secure backbone LLM sharing across isolated LoRA functions to reduce redundancy. We design a pre-loading method that pre-loads comprehensive LoRA artifacts to minimize cold-start latency. Furthermore, ServerlessLoRA employs contention aware batching and offloading to mitigate GPU resource conflicts during bursty workloads. Experiment on industrial workloads demonstrates that ServerlessLoRA reduces TTFT by up to 86% and cuts monetary costs by up to 89% compared to state-of-the-art LLM inference solutions.

cs.LG↗

arXiv2020

An Approach for GCI Fusion With Labeled Multitarget Densities

This paper addresses the Generalized Covariance Intersection (GCI) fusion method for labeled random finite sets. We propose a joint label space for the support of fused labeled random finite sets to represent the label association between different agents, avoiding the label consistency condition for the label-wise GCI fusion algorithm. Specifically, we devise the joint label space by the direct product of all label spaces for each agent. Then we apply the GCI fusion method to obtain the joint labeled multi-target density. The joint labeled RFS is then marginalized into a general labeled RFS, providing that each target is represented by a single Bernoulli component with a unique label. The joint labeled GCI (JL-GCI) for fusing LMB RFSs from different agents is demonstrated. We also propose the simplified JL-GCI method given the assumption that targets are well-separated in the scenario. The simulation result presents the effectiveness of label inconsistency and excellent performance in challenging tracking scenarios.

eess.SP↗

arXiv2010

How to Identify and Separate Bright Galaxy Clusters from the Low-frequency Radio Sky?

In this work we simulate the $50-200$ MHz radio sky that is constrained in the field of view ($5^{\circ}$ radius) of the 21 Centimeter Array (21CMA), by carrying out Monte-Carlo simulations to model redshifted cosmological reionization signals and strong contaminating foregrounds, including emissions from our Galaxy, galaxy clusters, and extragalactic point sources. As an improvement of previous works, we consider in detail not only random variations of morphological and spectroscopic parameters within the ranges allowed by multi-band observations, but also evolution of radio halos in galaxy clusters, assuming that relativistic electrons are re-accelerated in the ICM in merger events and lose energy via both synchrotron emission and inverse Compton scattering with CMB photons. By introducing a new approach designed on the basis of independent component analysis (ICA) and wavelet detection algorithm, we prove that, with a cumulative observation of one month with the 21CMA array, about $80\%$ of galaxy clusters with central brightness temperatures of $> 10~{\rm K}$ at 65 MHz can be safely identified and separated from the overwhelmingly bright foreground. We find that the morphological and spectroscopic distortions are extremely small as compared to the input simulated clusters, and the reduced $χ^2$ of brightness temperature profiles and spectra are controlled to be $\lesssim 0.5$ and $\lesssim 1.3$, respectively. These results robustly indicate that in the near future a sample of dozens of bright galaxy clusters will be disentangled from the foreground in 21CMA observations, the study of which will greatly improve our knowledge about cluster merger rates, electron acceleration mechanisms in cluster radio halos, and magnetic field in the ICM.

astro-ph.CO↗