papers

Publications (29)

math.DS2023

Mean ergodic theorems in and ,

el Houcein el Abdalaoui, Michael Lin

Let be the Koopman operator of a measure preserving transformation of a probability space . We study the convergence properties of the averages $M_nf:=\frac1n\s…

math.SP2026

Spectra of averages of unitary representations of LCA groups

Guy Cohen, Michael Lin

The paper studies the spectrum of operators obtained by averaging unitary representations of locally compact Abelian groups with respect to a probability measure, establishing incl…

#unitary representations#locally compact abelian groups#spectral theory#averaging operators
stat.AP2021

Stabilizing a Queue Subject to Action-Dependent Server Performance

Michael Lin, Richard J. La, Nuno C. Martins

We consider a discrete-time system comprising a first-come-first-served queue, a non-preemptive server, and a scheduler that governs the assignment of tasks in the queue to the ser…

cs.CL2019

SuperChat: Dialogue Generation by Transfer Learning from Vision to Language using Two-dimensional Word Embedding and Pretrained ImageNet CNN Models

Baohua Sun, Lin Yang, Michael Lin +4

The recent work of Super Characters method using two-dimensional word embedding achieved state-of-the-art results in text classification tasks, showcasing the promise of this new a…

math.DG2025

The Positive Mass Theorem for Creased Initial Data

Demetre Kazaras, Marcus Khuri, Michael Lin

We establish a spacetime positive mass theorem and rigidity statement for asymptotically flat spin initial data sets with a codimension one singularity controlled by a matching Bar…

cs.CV2019

SuperTML: Two-Dimensional Word Embedding for the Precognition on Structured Tabular Data

Baohua Sun, Lin Yang, Wenhan Zhang +4

Tabular data is the most commonly used form of data in industry. Gradient Boosting Trees, Support Vector Machine, Random Forest, and Logistic Regression are typically used for clas…

cs.RO2026

AGILE: A Comprehensive Workflow for Humanoid Loco-Manipulation Learning

Huihua Zhao, Rafael Cathomen, Lionel Gulich +6

Recent advances in reinforcement learning (RL) have enabled impressive humanoid behaviors in simulation, yet transferring these results to new robots remains challenging. In many r…

math.FA2020

Reflexive Banach spaces with all power-bounded operators almost periodic

Michael Lin

We analyze the ergodic properties of power-bounded operators on a reflexive Banach space of the form "scalar plus compact-power", and show that they are almost periodic (all the or…

cs.LG2025

FSL-SAGE: Accelerating Federated Split Learning via Smashed Activation Gradient Estimation

Srijith Nair, Michael Lin, Peizhong Ju +3

Collaborative training methods like Federated Learning (FL) and Split Learning (SL) enable distributed machine learning without sharing raw data. However, FL assumes clients can tr…

cs.CV2020

SuperOCR: A Conversion from Optical Character Recognition to Image Captioning

Baohua Sun, Michael Lin, Hao Sha +1

Optical Character Recognition (OCR) has many real world applications. The existing methods normally detect where the characters are, and then recognize the character for each detec…

math.DS2019

Twisty Takens: A Geometric Characterization of Good Observations on Dense Trajectories

Boyan Xu, Christopher J. Tralie, Alice Antia +2

In nonlinear time series analysis and dynamical systems theory, Takens' embedding theorem states that the sliding window embedding of a generic observation along trajectories in a…

cs.AI2025

WebGuard: Building a Generalizable Guardrail for Web Agents

Boyuan Zheng, Zeyi Liao, Scott Salisbury +8

The rapid development of autonomous web agents powered by Large Language Models (LLMs), while greatly elevating efficiency, exposes the frontier risk of taking unintended or harmfu…

cs.CL2020

Multi-modal Sentiment Analysis using Super Characters Method on Low-power CNN Accelerator Device

Baohua Sun, Lin Yang, Hao Sha +1

Recent years NLP research has witnessed the record-breaking accuracy improvement by DNN models. However, power consumption is one of the practical concerns for deploying NLP system…

math.FA2017

On modulated ergodic theorems

Tanja Eisner, Michael Lin

Let be a weakly almost periodic (WAP) linear operator on a Banach space . A sequence of scalars {\it modulates} on if $\frac1n\sum_{k=1}^n…

math.OC2021

Channels, Remote Estimation and Queueing Systems With A Utilization-Dependent Component: A Unifying Survey Of Recent Results

Varun Jog, Richard J. La, Michael Lin +1

In this article, we survey the main models, techniques, concepts, and results centered on the design and performance evaluation of engineered systems that rely on a utilization-dep…

cs.LG2026

Stabilizing Transformer Training Through Consensus

Shyam Venkatasubramanian, Sean Moushegian, Michael Lin +3

Standard attention-based transformers are known to exhibit instability under learning rate overspecification during training, particularly at high learning rates. While various met…

cs.CL2019

Squared English Word: A Method of Generating Glyph to Use Super Characters for Sentiment Analysis

Baohua Sun, Lin Yang, Catherine Chi +2

The Super Characters method addresses sentiment analysis problems by first converting the input text into images and then applying 2D-CNN models to classify the sentiment. It achie…

cs.RO2024

Tactile-Informed Action Primitives Mitigate Jamming in Dense Clutter

Dane Brouwer, Joshua Citron, Hojung Choi +4

It is difficult for robots to retrieve objects in densely cluttered lateral access scenes with movable objects as jamming against adjacent objects and walls can inhibit progress. W…

math.PR2026

Doeblin's condition, -mixing and spectra of convolution operators on the circle

Guy Cohen, Michael Lin

We study the asymptotic behavior of Markov operators defined by convolution with a probability measure on the unit circle . We prove that when is adapte…

q-bio.QM2021

Inferring, comparing and exploring ecological networks from time-series data through R packages constructnet, disgraph and dynet

Anshuman Swain, Travis Byrum, Zhaoyi Zhuang +3

Network inference is a major field of interest for the ecological community, especially in light of the high cost and difficulty of manual observation, and easy availability of rem…

math.DS2020

Joint and double coboundaries of commuting contractions

Guy Cohen, Michael Lin

Let and be commuting contractions on a Banach space . The elements of are called {\it double coboundaries}, and the elements of are ca…

cs.CL2019

SuperCaptioning: Image Captioning Using Two-dimensional Word Embedding

Baohua Sun, Lin Yang, Michael Lin +4

Language and vision are processed as two different modal in current work for image captioning. However, recent work on Super Characters method shows the effectiveness of two-dimens…

math.DS2023

Uniform ergodicity and the one-sided ergodic Hilbert transform

Guy Cohen, Michael Lin

Let be a bounded linear operator on a Banach space satisfying . We prove that is uniformly ergodic if and only if the one-sided ergodic Hilbert transfo…

math.DS2026

Operator ergodic theorems with Möbius "weights"

El Houcein El Abdalaoui, Michael Lin

Motivated by Sarnak's conjecture in topological dynamics for the Möbius function , we study, for a power-bounded on a Banach space , the weak convergence $$ (*) \qquad…

cs.AI2025

Mind2Web 2: Evaluating Agentic Search with Agent-as-a-Judge

Boyu Gou, Zanming Huang, Yuting Ning +23

Agentic search such as Deep Research systems-where agents autonomously browse the web, synthesize information, and return comprehensive citation-backed answers-represents a major s…

stat.AP2020

Queueing Subject To Action-Dependent Server Performance: Utilization Rate Reduction

Michael Lin, Nuno C. Martins, Richard J. La

We consider a discrete-time system comprising a first-come-first-served queue, a non-preemptive server, and a stationary non-work-conserving scheduler. New tasks enter the queue ac…

math.DS2020

Resolvent conditions and growth of powers of operators

Guy Cohen, Christophe Cuny, Tanja Eisner +1

Following Bermúdez et al. (ArXiv: 1706.03638v1), we study the rate of growth of the norms of the powers of a linear operator, under various resolvent conditions or Cesà ro bounded…

cs.CL2019

System Demo for Transfer Learning across Vision and Text using Domain Specific CNN Accelerator for On-Device NLP Applications

Baohua Sun, Lin Yang, Michael Lin +4

Power-efficient CNN Domain Specific Accelerator (CNN-DSA) chips are currently available for wide use in mobile devices. These chips are mainly used in computer vision applications.…

cs.DS2020

Miniature Robot Path Planning for Bridge Inspection: Min-Max Cycle Cover-Based Approach

Michael Lin, Richard J. La

We study the problem of planning the deployments of a group of mobile robots. While the problem and formulation can be used for many different problems, here we use a bridge inspec…