Zixiang Chen

[08/2026] Promoted to Senior Research Scientist at Salesforce AI Research.

[03/2026] Our paper on vector prompt interfaces was accepted to ICML 2026.

[12/2025] Attended NeurIPS 2025.

[09/2025] Released Enterprise Deep Research, an open-source multi-agent framework for enterprise analytics.[code] 

[07/2025] Joined Salesforce AI Research.

[06/2025] Graduated from UCLA.

[05/2025] Our work on feature learning and global convergence of deep neural networks was accepted to ICML 2025.

[01/2025] Our discrete diffusion analysis was accepted to ICLR 2025.

[01/2025] Our work on multitask representation learning was accepted to AISTATS 2025.

[12/2024] Attended NeurIPS 2024 in Vancouver.

[11/2024] Gave a lightning talk at Google’s Theory and Practice of Foundation Models Workshop.

[09/2024] Three papers were accepted to NeurIPS 2024.

[06/2024] Received the UCLA Dissertation Year Fellowship.

[05/2024] SPIN was accepted to ICML 2024.

[01/2024] Two papers on multimodal learning and in-context learning were accepted to ICLR 2024.

[01/2024] Released Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models (SPIN).

[12/2023] Attended NeurIPS in New Orleans.

[09/2023] Two papers were accepted to NeurIPS 2023.

[07/2023] Attended ICML in Hawaii.

[05/2023] Two papers were accepted to ICML 2023.

[01/2023] Three papers were accepted to ICLR 2023.

[12/2022] Attended my first in-person conference, NeurIPS.

[09/2022] Two papers were accepted to NeurIPS 2022.

[08/2022] Gave a talk on Towards Understanding Mixture of Experts in Deep Learning at the TTIC Chicago Summer Workshop on Representation Learning Theory.

[06/2022] Passed my oral qualifying exam.

[05/2022] Gave a talk on Benign Overfitting in Two-layer Convolutional Neural Networks at the Math Machine Learning Seminar hosted by MPI MiS and UCLA.