Zhixin Zeng 曾智鑫

Independent Researcher

I work on machine learning and reinforcement learning. My current research is on agents for video editing: giving an agent the checks, feedback, and editable history that coding agents already get from compilers and version control. Earlier, I worked on inverse reinforcement learning for urban safety perception and on fake news detection. I hold an M.S. in Computer Science from Northeastern University and a B.S. from Sichuan University.

Publications

2026
Cassette-Editor: An Agent Harness for Editing Video the Way You Do

Zhixin Zeng, Jiahao Zhong

Preprint, arXiv:2610.37690 · Website · GitHub
2024
Fake News Detection Based on Knowledge-Guided Semantic Analysis

Wenbin Zhao, Peisong He, Zhixin Zeng, Xiong Xu

Electronics 13(2), 259
2022
Evaluating the Perceived Safety of Urban City via Maximum Entropy Deep Inverse Reinforcement Learning

Yaxuan Wang, Zhixin Zeng, Qijun Zhao

Asian Conference on Machine Learning (ACML 2022), PMLR 189:1085–1100 · arXiv
2022
A Complete Reinforcement-Learning-Based Framework for Urban-Safety Perception

Yaxuan Wang*, Zhixin Zeng*, Qiushan Li, Yingrui Deng

ISPRS International Journal of Geo-Information 11(9), 465 · *Equal contribution

Projects

Cassette

A video editor in which people and agents make edits through the same path, available as a browser editor and a macOS app. Co-founder and primary developer.
oh-my-cassette

An open-source plugin that connects coding agents, including Claude Code and Codex, to the Cassette editor.

Visitors

World map of recent visitor locations