Tong Xie

hi there! 👋

I am a second-year Ph.D. student in Computer Science at UCLA, fortunate to be advised by Prof. Cho-Jui Hsieh.

I am broadly interested in post-training of large language models. Currently, I am working on supervised fine-tuning (SFT), reinforcement learning (RL), and reward modeling, to improve reasoning capabilities and encourage stronger generalization.

Feel free to connect with me and explore opportunities, collaborations, and exciting ventures together!

News

Sep. 2026 Arena-T2I Hard is accepted to NeurIPS 2026 as a Spotlight.
Jun. 2026 Target-SFT is now on arXiv. Check out the project page!
May 2026 NormBT is accepted to ICML 2026.
Sep. 2025 I started my PhD at UCLA!
Jun. 2024 Diffusion-ReTrac is accepted to TMLR 2024.
Jun. 2024 I joined QSG for the summer— an awesome buy-side quant group at RBC.
Jun. 2023 I joined UCLA's undergraduate research program (SURP) for the summer. Check out the poster!

Selected Publications

Preprint

A Unifying Lens on Supervised Fine-Tuning Through Target Distribution Design

Tong Xie, Yuanhao Ban, Yunqi Hong, Sohyun An, Yihang Chen, Cho-Jui Hsieh

Preprint, 2026.

PDF Code Project Page

NeurIPS 2026

Arena-T2I Hard: Benchmarking and Improving Faithfulness with Dependency-Aware Checklist

Yuanhao Ban, Tong Xie, Sohyun An, Yunqi Hong, Evan Frick, I-Hung Hsu, Wei-Lin Chiang, Ion Stoica, Cho-Jui Hsieh

Neural Information Processing Systems (NeurIPS), 2026.

PDF Code Dataset Project Page Spotlight

ICML 2026

When Distance Distracts: Representation Distance Bias in BT-Loss for Reward Models

Tong Xie, Andrew Bai, Yuanhao Ban, Yunqi Hong, Haoyu Li, Cho-Jui Hsieh

International Conference on Machine Learning (ICML), 2026.

PDF Code

TMLR 2024

Data Attribution for Diffusion Models: Timestep-induced Bias in Influence Estimation

Tong Xie*, Haoyu Li*, Andrew Bai, Cho-Jui Hsieh

Transactions on Machine Learning Research (TMLR), 2024.

PDF Code

Last updated: October 2026.
Powered by Jekyll and Minimal Light theme.