ACL 2026 · Main Conference
LLM Agents · NLP · Human-centered AI
Building agents that learn from experience and communicate with people.
I am Ziyang Zhou (Deric). I recently completed my B.Sc. in Information and Computer Science at Xi'an Jiaotong-Liverpool University. My research focuses on LLM agents, natural language processing, and audio & speech.
I am particularly interested in how multiple agents can reason together, reuse experience, and produce more reliable and empathetic responses. I will begin graduate study in Fall 2026.
ICASSP 2026
Following the TRACE accepted to ICASSP 2026.
View publication ↗ICML Workshop on CFAgentic
Agent KB received the Best Paper Runner-up Award.
Read paper ↗01 / Research
Selected publications
Work on agentic reasoning, affective language understanding, and multi-agent collaboration. Full list on Scholar ↗
02 / Practice
Research experience
Selected collaborations across agent systems, language computing, and audio intelligence.
Human-Centered Language Computing Group
XJTLU
Research on human-centered language technologies, affective understanding, and multi-agent reasoning.
Gerstein Lab
Yale University
Research collaboration on LLM agents and experience-driven intelligent systems.
Lab ↗
Spiro — Axiomsten
Agent Algorithm Intern
Worked on agent algorithms, reasoning workflows, and practical agent system development.
AIC-XJTLU
Audio Intelligence Research
Research on accent recognition and acoustic scene classification using HuBERT and device-aware audio modeling.
OPPO AI Agent Team
Research Collaborator
Explored code agents and agentic approaches to software development.
03 / Background
Education & recognition
A foundation in computer science, research, and collaborative problem solving.
Education
Xi'an Jiaotong-Liverpool University
B.Sc. in Information and Computer Science
Best Paper Runner-up
ICML Workshop on CFAgentic · Agent KB
National Third Prize
Lanqiao Java Programming Competition
S Prize
Mathematical Contest in Modeling
04 / Beyond research
Technology feels most meaningful when it stays close to human expression.
My interests in cinema, music, traditional calligraphy, and film sound shape how I think about intelligence: not only as performance, but also as attention, interpretation, and communication.
Let's talk about research ↗