Tian Xia · Language models and post-training

Tian Xia, Applied Scientist at MicrosoftTian Xia / The StudioPepper is my cat. The illustrated resident of this desk is Pepper, too. The desk itself is drawn with JavaScript; without it, this is the reading view.ResearchEvery paper, newest first. A title that is a link opens its interactive page; the links after it open the paper, as a file where this site hosts it, and at its publisher.

AACL-IJCNLP 2026, to appear Three Numbers and a Precondition Read the paper (PDF) (opens in a new tab)

ACM CAIS 2026 SAPO: Secure Automated Prompt Optimization via Multi-Agent Collaboration Read the paper (opens in a new tab)

EMNLP 2025 How to Fine-Tune Safely on a Budget: Model Adaptation Using Minimal Resources Read the paper (PDF) (opens in a new tab) At the ACL Anthology (opens in a new tab)

NeurIPS 2025 workshops High-Fidelity Synthetic ECG Generation via Mel-Spectrogram Informed Diffusion Training Read the paper (PDF, 4.5 MB) (opens in a new tab) On arXiv (opens in a new tab)

ICLR 2025 GotenNet: Rethinking Efficient 3D Equivariant Graph Neural Networks Read the paper (PDF) (opens in a new tab) At OpenReview (opens in a new tab)

IEEE BigData 2024 SO(3) Equivariant Framework for Spatial Networks Read the paper (opens in a new tab)

IEEE BigData 2024 SE(3) Equivariant Neural Network for 3D Graphs Read the paper (opens in a new tab)

NeurIPS 2023 SaVeNet: A Scalable Vector Network for Enhanced Molecular Representation Learning Read the paper (PDF) (opens in a new tab) At NeurIPS (opens in a new tab)

AAAI 2023 Constrained Market Share Maximization by Signal-Guided Optimization Read the paper (opens in a new tab)

SDM 2023 Extrinsic-Intrinsic Representation Learning Framework for Drug Discovery Read the paper (opens in a new tab)

EMNLP 2022 A Localized Geometric Method to Match Knowledge in Low-dimensional Hyperbolic Space Read the paper (PDF) (opens in a new tab) At the ACL Anthology (opens in a new tab)

KDD 2022 GBPNet: Universal Geometric Representation Learning on Protein Structures Read the paper (opens in a new tab)

IEEE BigData 2022 APIP: Attention-based Protein Representation Learning for Protein-Ligand Interface Prediction Read the paper (opens in a new tab)

KDD 2021 Geometric Graph Representation Learning on Protein Structure Prediction Read the paper (opens in a new tab)

The research indexElsewhere

LinkedIn (opens in a new tab) My professional profile