Multiple papers accepted at top-tier venues including NeurIPS 2025 (2 papers), ICCV 2025, ICLR 2025 (2 papers), TMLR (multiple), ECCV 2024, PR 2024, etc.
Nominated as Area Chair for ICLR 2026 by Program Chairs (Aug 2025)
Guest Editor for the Entropy Special Issue “Rethinking Representation Learning in the Age of Large Models” (2025)
Prolific contributions in causal inference, domain adaptation, representation learning, interpretability of LLMs, and multimodal learning
Key focus on identifiability of latent causal models, data diversity, and representation alignment
Background
Research Fellow at the Australian Institute for Machine Learning (AIML), The University of Adelaide
Collaborating with Prof. Qinfeng (Javen) Shi
Research mission: Building Kant’s Bridge through the lens of data diversity
Core question: How to align machine-learned representations with the underlying knowledge of data, analogous to the philosophical distinction between appearance and essence
Short-term objectives: Enhancing interpretability, controllability, and reasoning in large language models; exploring text diversity to uncover latent patterns in unstructured data; developing foundational theories linking latent variable models and representation learning via data diversity