arXiv subjects

Zitong Zhao

Publications and source records attributed to Zitong Zhao.

9 recordsLinked to original sources

arXiv2026

A Multimodal Pre-trained Network for Integrated EEG-Video Seizure Detection

Reliable seizure detection in mouse models is essential for preclinical epilepsy research, yet manual review of synchronized video-EEG recordings is labor-intensive and single-modality systems fail for complementary reasons: video-based methods are easily confounded by benign behaviors, whereas EEG-based methods are vulnerable to ictal motion artifacts. We present EEGVFusion, a multimodal framework that combines self-supervised EEG representation learning, spatio-temporal video encoding, optimal-transport alignment, and bidirectional cross-attention to integrate neural and behavioral evidence. We also curate an expert-annotated dataset of synchronized EEG and video recordings comprising 93 sessions from 15 mice for training and evaluation. In the random-session split, EEGVFusion achieved a Balanced Accuracy of 0.9957 with perfect event sensitivity and an Event FAR of 0.6250 FP/h, indicating strong seizure detection performance with a low false-alarm burden. In a single held-out-subject evaluation with Subject 110 reserved for testing, EEGVFusion achieved a Balanced Accuracy of 0.9718 and reduced Event FAR from 2.7250 FP/h for the EEG-only counterpart to 0.4833 FP/h while preserving perfect event sensitivity. Targeted ablations further showed that EEG pre-training and OT alignment help reduce false alarms while preserving event sensitivity.

cs.CV↗

arXiv2026

Necessary conditions for existence of tensor invariants for general nonlinear dynamical systems

The integrability has been playing an essential role in the field of differential equations. This property may better help us obtain the topological structure and even the global dynamics for the considered system. A system is called integrable if it has a number of tensor invariants, which can comprehensively define the integrability problem. In this paper, we give necessary conditions for existence of tensor invariants for general nonlinear systems, especially semi-quasihomogeneous systems. Our results may be viewed as a generalization of Poincaré and Kozlov's work.

math.DS↗

arXiv2025

Uncovering Scaling Laws for Large Language Models via Inverse Problems

Large Language Models (LLMs) are large-scale pretrained models that have achieved remarkable success across diverse domains. These successes have been driven by unprecedented complexity and scale in both data and computations. However, due to the high costs of training such models, brute-force trial-and-error approaches to improve LLMs are not feasible. Inspired by the success of inverse problems in uncovering fundamental scientific laws, this position paper advocates that inverse problems can also efficiently uncover scaling laws that guide the building of LLMs to achieve the desirable performance with significantly better cost-effectiveness.

cs.LG↗

arXiv2025

Fermi Level and Light Driven Defect Generation in Silicon Solar Cells

Hydrogenated amorphous silicon (a-Si:H) has had a long standing role as a passivating dielectric for c-Si, often utilized in the early development of ICs and more recently for Si solar cells. Although it has been studied for more than 60 years, several questions about the material properties remain open, including light-induced degradation and the Fermi level dependence on the mobility of hydrogen. Here we study the origin of these phenomenon using electronic structure based calculations. First, we use density functional theory (DFT) and the nudged elastic band (NEB) method to examine defect generation via Si-H bond breaking in p-type, intrinsic, and n-type a-Si:H. We find that the energy barrier controlling this defect generation, shows the same asymmetric reduction of $\sim 0.3$ eV for p-type and $\sim 0.1$ eV for n-type, observed in experimental studies. We then develop a model based on the local Coulomb interactions at the transition state, which provides compelling evidence that the asymmetry results from the emergence of a high energy donor state created by the interstitial H. Finally, we repeat our Si-H bond breaking analysis, combining NEB with constrained density functional perturbation theory (c-DFPT) to simulate defect generation dynamics in illuminated a-Si:H. Here we find that e-p pair results in a combined effect that reduces the barrier by $\sim 0.4$, in close agreement with experimental observations.

cond-mat.mtrl-sci↗

arXiv2024

Source Attribution for Large Language Model-Generated Data

The impressive performances of Large Language Models (LLMs) and their immense potential for commercialization have given rise to serious concerns over the Intellectual Property (IP) of their training data. In particular, the synthetic texts generated by LLMs may infringe the IP of the data being used to train the LLMs. To this end, it is imperative to be able to perform source attribution by identifying the data provider who contributed to the generation of a synthetic text by an LLM. In this paper, we show that this problem can be tackled by watermarking, i.e., by enabling an LLM to generate synthetic texts with embedded watermarks that contain information about their source(s). We identify the key properties of such watermarking frameworks (e.g., source attribution accuracy, robustness against adversaries), and propose a source attribution framework that satisfies these key properties due to our algorithmic designs. Our framework enables an LLM to learn an accurate mapping from the generated texts to data providers, which sets the foundation for effective source attribution. Extensive empirical evaluations show that our framework achieves effective source attribution.

cs.LG↗

arXiv2023

Hydrogen-induced degradation dynamics in silicon heterojunction solar cells via machine learning

Among silicon-based solar cells, heterojunction cells hold the world efficiency record. However, their market acceptance is hindered by an initial 0.5\% per year degradation of their open circuit voltage which doubles the overall cell degradation rate. Here, we study the performance degradation of crystalline-Si/amorphous-Si:H heterojunction stacks. First, we experimentally measure the interface defect density over a year, the primary driver of the degradation. Second, we develop SolDeg, a multiscale, hierarchical simulator to analyze this degradation by combining Machine Learning, Molecular Dynamics, Density Functional Theory, and Nudged Elastic Band methods with analytical modeling. We discover that the chemical potential for mobile hydrogen develops a gradient, forcing the hydrogen to drift from the interface, leaving behind recombination-active defects. We find quantitative correspondence between the calculated and experimentally determined defect generation dynamics. Finally, we propose a reversed Si-density gradient architecture for the amorphous-Si:H layer that promises to reduce the initial open circuit voltage degradation from 0.5\% per year to 0.1\% per year.

cond-mat.mtrl-sci↗

arXiv2021

Assessing Attendance by Peer Information

Attendance rate is an important indicator of students' study motivation, behavior and Psychological status; However, the heterogeneous nature of student attendance rates due to the course registration difference or the online/offline difference in a blended learning environment makes it challenging to compare attendance rates. In this paper, we propose a novel method called Relative Attendance Index (RAI) to measure attendance rates, which reflects students' efforts on attending courses. While traditional attendance focuses on the record of a single person or course, relative attendance emphasizes peer attendance information of relevant individuals or courses, making the comparisons of attendance more justified. Experimental results on real-life data show that RAI can indeed better reflect student engagement.

cs.SI↗

arXiv2020

Anti-Parity-Time Symmetry in Passive Nanophotonics

Parity-time (PT) symmetry in non-Hermitian optical systems promises distinct optical effects and applications not found in conservative optics. Its counterpart, anti-PT symmetry, subscribes another class of intriguing optical phenomena and implies complementary techniques for exotic light manipulation. Despite exciting progress, so far anti-PT symmetry has only been realized in bulky systems or with optical gain. Here, we report an on-chip realization of non-Hermitian optics with anti-PT symmetry, by using a fully-passive, nanophotonic platform consisting of three evanescently coupled waveguides. By depositing a metal film on the center waveguide to introduce strong loss, an anti-PT system is realized. Using microheaters to tune the waveguides' refractive indices, striking behaviors are observed such as equal power splitting, synchronized amplitude modulation, phase-controlled dissipation, and transition from anti-PT symmetry to its broken phase. Our results highlight exotic anti-Hermitian nanophotonics to be consolidated with conventional circuits on the same chip, whereby valuable chip devices can be created for quantum optics studies and scalable information processing.

physics.optics↗

arXiv2019

Application of a Modified Harmony Search Algorithm in the Optimal Arrangement of a Novel Three Dimensional Multiphase Flow Imaging Device

Gas-liquid two-phase flow is a typical flow, and bubble characteristic measurement is of great importance to study the flow mechanism and guide the practical fluid mechanical engineering. In this paper, a novel three dimensional (3D) multiphase flow imaging device was designed to measure the transparent object that has an opaque object in the center of the observed area. Its mathematical model was built and the constraints were defined based on the geometrical relationship and design requirements. A modified harmony search (HS) algorithm was integrated and applied to optimize the arrangement of the single-camera-multi-mirror device. As a case study, the 3D multiphase flow imaging method was applied in the the 3D reconstruction of the cavitation bubble cluster inside a water hydraulic valve. The statistics of the Pareto data shows the good performance of the modified HS algorithm. And the cavitation experimental results shows that the method is valid, and the cavitation bubble cluster can be reconstructed with quite high precision.

eess.IV↗