Source author record

Yongsheng Huang

Yongsheng Huang appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Machine Learning Artificial Intelligence hep-ex hep-ph physics.acc-ph physics.plasm-ph

Catalog footprint

What is connected

4works

6topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

Not All Timesteps Matter Equally: Selective Alignment Knowledge Distillation for Spiking Neural Networks

Spiking neural networks (SNNs), which are brain-inspired and spike-driven, achieve high energy efficiency. However, a performance gap between SNNs and artificial neural networks (ANNs) still remains. Knowledge distillation (KD) is commonly adopted to improve SNN performance, but existing methods typically enforce uniform alignment across all timesteps, either from a teacher network or through inter-temporal self-distillation, implicitly assuming that per-timestep predictions should be treated equally. In practice, SNN predictions vary and evolve over time, and intermediate timesteps need not all be individually correct even when the final aggregated output is correct. Under such conditions, effective distillation should not force every timestep toward the same supervision target, but instead provide corrective guidance to erroneous timesteps while preserving useful temporal dynamics. To address this issue, we propose Selective Alignment Knowledge Distillation (SeAl-KD), which selectively aligns class-level and temporal knowledge by equalizing competing logits at erroneous timesteps and reweighting temporal alignment based on confidence and inter-timestep similarity. Extensive experiments on static image and neuromorphic event-based datasets demonstrate consistent improvements over existing distillation methods. The code is available at https://github.com/KaiSUN1/SeAl

preprint2026arXiv

We Need a More Robust Classifier: Dual Causal Learning Empowers Domain-Incremental Time Series Classification

The World Wide Web thrives on intelligent services that rely on accurate time series classification, which has recently witnessed significant progress driven by advances in deep learning. However, existing studies face challenges in domain incremental learning. In this paper, we propose a lightweight and robust dual-causal disentanglement framework (DualCD) to enhance the robustness of models under domain incremental scenarios, which can be seamlessly integrated into time series classification models. Specifically, DualCD first introduces a temporal feature disentanglement module to capture class-causal features and spurious features. The causal features can offer sufficient predictive power to support the classifier in domain incremental learning settings. To accurately capture these causal features, we further design a dual-causal intervention mechanism to eliminate the influence of both intra-class and inter-class confounding features. This mechanism constructs variant samples by combining the current class's causal features with intra-class spurious features and with causal features from other classes. The causal intervention loss encourages the model to accurately predict the labels of these variant samples based solely on the causal features. Extensive experiments on multiple datasets and models demonstrate that DualCD effectively improves performance in domain incremental scenarios. We summarize our rich experiments into a comprehensive benchmark to facilitate research in domain incremental time series classification.

preprint2022arXiv

Light Dark Matter Axion Detection with Static Electric Field

We explore the axionic dark matter search sensitivity with a narrow-band detection scheme aiming at the axion-photon conversion by the static electric field inside a cylindrical capacitor. An alternating magnetic field signal is induced by effective currents as the axion dark matter flows perpendicularly through the electric field. At low axion masses, like in a KKLT scenario, front-end narrow band filtering is provided by using LC resonance with a high $Q$ factor, which enhances the detectability of the tiny magnetic field signal and also leads to a thermal noise as the major background that can be reduced at cryogenic conditions. We demonstrate that high $g_{aγ}$ sensitivity can be achieved by using a strong electric field. The QCD axion theoretical parameter space can be reached with high $E\sim$ GVm$^{-1}$ field strength. Using the static electric field scheme essentially avoids exposing the sensitive superconducting pickup to an applied laboratory magnetic field.

preprint2011arXiv

Double-relativistic-electron-shell laser proton acceleration

A new laser-proton acceleration structure combined by two relativistic electron shells, a suprathermal electron shell and a thermal electron cloud is proposed for $a\gtrapprox80σ_0$, where a is the normalized laser field and $σ_0$ is the normalized plasma surface density. In the new region, a uniform energy distribution of several GeV and a monoenergetic hundreds-of-MeV proton beam have been obtained for $a=39.5$. The first relativistic electron shell maintains opaque for incident laser pulse in the whole process. A monoenergetic electron beam has been generated with energy hundreds of MeV and charge of hundreds of pC. It is proposed a stirring solution for relativistic laser-particle acceleration.