Research

I work on machine learning systems that are robust, efficient, and useful across structured and unstructured data. Current interests include foundation models, retrieval, evaluation, and outlier detection.

Publications

LLM Evaluation, Pre-/Post-Training, and Foundation Models

2026

From Zero to Hero: Advancing Zero-Shot Foundation Models for Tabular Outlier Detection

X. Ding, H. Wen, S. Klutterman, L. Akoglu

ICML

2026

Hierarchical Token Prepending: Enhancing Information Flow in Decoder-based LLM Embeddings

X. Ding*, X. Huang*, M. Ju, L. Collins, Y. Liu, L. Akoglu, N. Shah, T. Zhao

ACL Main Track

Oral presentation

2026

Threshold Differential Attention for Sink-Free Ultra-Sparse, and Non-Dispersive Language Modeling

X. Huang*, X. Ding*, M. Ju, Y. Liu, N. Shah, T. Zhao

ACL Main Track

Oral presentation

2025

Firm or Fickle? Evaluating Large Language Models Consistency in Sequential Interactions

Y. Li, Y. Miao, X. Ding, R. Krishnan, R. Padman

ACL Findings

2026

Toward Privileged Foundation Models: LUPI for Accelerated and Improved Learning

X. Ding, L. Akoglu

Under Review

2025

DELPHYNE: A Pre-Trained Model for Financial Time Series

X. Ding, A. Mittal, A. Gopal

NeurIPS GenAI in Finance Workshop

Runner-up Best Paper Award

2024

Pard: Permutation-Invariant Autoregressive Diffusion for Graph Generation

L. Zhao, X. Ding, L. Akoglu

NeurIPS

2025

Improving and Unifying Discrete & Continuous-time Discrete Denoising Diffusion

L. Zhao*, X. Ding*, L. Yu, L. Akoglu

JMLR

Outlier Detection, Optimization, and Data-Centric ML

2026

MacrOData: New Benchmarks of Thousands of Datasets for Tabular Outlier Detection

X. Ding, S. Klutterman, H. Wen, Y. Chen, L. Akoglu

KDD

2024

Outlier Detection Bias Busted: Understanding Sources of Algorithmic Bias through Data-centric Factors

X. Ding, R. Xi, L. Akoglu

AAAI (AI, Ethics, and Society)

2025

MetaOOD: Automatic Selection of OOD Detection Models

Y. Qin, Y. Zhang, Y. Nian, X. Ding, Y. Zhao

ICLR

2024

Fast Unsupervised Deep Outlier Model Selection with Hypernetworks

X. Ding, Y. Zhao, L. Akoglu

KDD

2024

PINNsFormer: A Transformer-Based Framework for Physics-Informed Neural Networks

Z. Zhao, X. Ding, Y. Zhao, B. A. Prakash

ICLR

2023

From Explanation to Action: An End-to-End Human-in-the-loop Framework for Anomaly Reasoning and Management

X. Ding, N. Seleznev, S. Kumar, C. B. Bruss, L. Akoglu

ACM ICAIF

Oral presentation

2022

Hyperparameter Sensitivity in Deep Outlier Detection: Analysis and a Scalable Hyper-Ensemble Solution

X. Ding, L. Zhao, L. Akoglu

NeurIPS

2024

PyGOD: A Python Library for Graph Outlier Detection

K. Liu, Y. Dou, X. Ding, X. Hu, R. Zhang, H. Peng, L. Sun, P. Yu

JMLR