Changling Li

I'm a computer science master student at ETH Zurich majoring in machine intelligence and minoring in theoretical computer science. My research interests center on the dynamic interplay between artificial intelligence and society, with a particular focus on reinforcement learning, multi-agent systems, and AI ethics. This passion for cooperation and ethical considerations is deeply rooted in my personal experiences and identity, especially during my time at United World College where I witnessed the power of communication and mutual respect in building understanding and unity despite massive culture differences. I believe that developing interactions among AIs, and between AIs and humans requires the same foundational principles with both technical innovation and philosophical advancement.

I am currently working with Hui Zhang in collaboration with Prof.Mirko Meboldt at ETH Zurich and with Zhang-Wei Hong in collaboration with Prof.Pulkit Agrawal at MIT and Prof.Joni Pajarinen at Aalto University. Previously, I obtained my B.A. degrees in computer science and physics with a concentration in astrophysics at Colby College. I was fortunate to work with Prof.Ying Li for my honor thesis which led me to discover my research interest in decision-making in multi-agent systems.

Outside research, I enjoy watching films and sometimes write short critiques. I also love playing squash and rowing. Before I got into computer science, I was an art student.

Email  /  CV  /  Scholar  /  Twitter  /  Github

profile photo

Publication

ROER: Regularized Optimal Experience Replay
Changling Li, Zhang-Wei Hong, Pulkit Agrawal, Divyansh Garg, Joni Pajarinen
RLC, 2024
Code / Poster / arXiv

By formulating experience prioritization as an occupancy optimization problem, we show that the form of TD-error based prioritization is closely associated with the form of the loss function and thus, we propose a new formulation of prioritization using KL-divergence.

Energy-Aware Multi-Agent Reinforcement Learning for Collaborative Execution in Mission-Oriented Drone Networks
Ying Li, Changling Li, Jiyao Chen, Christine Roinou
ICCCN, 2022

Creating multi-agent reinforcement learning framework for mission-oriented drone networks to enable energy-efficient task planning and execution.

Teaching

Teaching Assistant, Department of Computer Science, Colby College

CS 353 Interactive System

CS 251 Data Analysis and Visualization

CS 231 Data Structure and Algorithm

CS 152 Computational Thinking: Science

CS 151 Computational Thinking: Visual Media

Teaching Assistant, Department of Physics and Astronomy, Colby College

PH 241 Modern Physics I

PH 242 Modern Physics II


template from jonbarron