Foundations of modern machine learning
We build mathematical frameworks to better understand the success of existing ML algorithms, and devise methods to overcome their limitations.
Our current focus is on foundation models (including large language models and vision-language models) studied across four directions.
Internal representations & mechanisms of foundation models
We study what foundation models encode and how they compute it internally. This spans the geometry of learned representations, the theory of contrastive and self-supervised learning, compositional reasoning in vision-language models, and the mechanisms by which transformers process information. The goal is a mathematical account of why these models represent the world the way they do.
- Enhancing Compositional Reasoning in CLIP via Reconstruction and Alignment of Text Descriptions NeurIPS · 2025
- Soft Task-Aware Routing of Experts for Equivariant Representation Learning NeurIPS · 2025
- On the Similarities of Embeddings in Contrastive Learning ICML · 2025
- A Theoretical Framework for Preventing Class Collapse in Supervised Contrastive Learning AISTATS · 2025
- Mini-Batch Optimization of Contrastive Loss TMLR · 2024
- Analysis of Using Sigmoid Loss for Contrastive Learning AISTATS · 2024
- Looped Transformers as Programmable Computers ICML · 2023
- Utilizing Language-Image Pretraining for Efficient and Robust Bilingual Word Alignment Findings of EMNLP · 2022
Training & adaptation of foundation models
We design principled ways to train, fine-tune, and continually adapt foundation models. Our work covers fine-tuning that preserves in-context learning, the memorization capacity of fine-tuning, continual learning under distribution shift, and language-interfaced fine-tuning for non-language tasks. We aim for adaptation methods backed by provable guarantees rather than purely empirical recipes.
- Fine-Tuning Without Forgetting In-Context Learning: A Theoretical Analysis of Linear Attention Models ICML · 2026
- Measuring Representational Shifts in Continual Learning: A Linear Transformation Perspective ICML · 2025
- Memorization Capacity for Additive Fine-Tuning with Small ReLU Networks UAI · 2024
- LIFT: Language-Interfaced Fine-Tuning for Non-Language Machine Learning Tasks NeurIPS · 2022
Trustworthy & efficient foundation models
We make foundation models reliable and affordable to deploy. The aim is trustworthy behavior at lower cost, without giving up capability.
- How to Correctly Report LLM-as-a-Judge Evaluations ICML · 2026
- Can We Find Strong Lottery Tickets in Generative Models? AAAI · 2023
- Rare Gems: Finding Lottery Tickets at Initialization NeurIPS · 2022
- Finding Everything within Random Binary Networks AISTATS · 2022
Applications of foundation models
We bring foundation models into real-world domains alongside domain experts. Current applications include healthcare and psychiatry (cognitive behavioral therapy alignment, psychiatric interviews, cognitive-distortion classification), finance (question answering and retrieval-augmented generation), and law (Korean legal QA). These deployments create real impact while surfacing new theoretical questions.
- Aligning Large Language Models for Cognitive Behavioral Therapy: A Proof-of-Concept Study Frontiers in Psychiatry · 2025
- FinDER: Financial Dataset for Question Answering and Retrieval-Augmented Generation ICLR Workshop · 2025
- Aligning Large Language Models for Enhancing Psychiatric Interviews through Symptom Delineation and Summarization JMIR Formative Research · 2024
- ERD: A Framework for Improving LLM Reasoning for Cognitive Distortion Classification NAACL Workshop · 2024
- Retrieval-based Evaluation for LLMs: A Case Study in Korean Legal QA EMNLP Workshop · 2023
Earlier research directions
Trustworthy ML
Robust, fair, and private machine learning.
Distributed ML
Distributed and federated learning over networks.
- Buffer-based Gradient Projection for Continual Federated Learning TMLR · 2025
- Election Coding for Distributed Learning: Protecting SignSGD against Byzantine Attacks NeurIPS · 2020
- Attack of the Tails: Yes, You Really Can Backdoor Federated Learning NeurIPS · 2020
- Hierarchical Coding for Distributed Computing IEEE ISIT · 2018
Information / Communication Theory
Coding theory for storage, computing, and wireless systems.
- Capacity of Clustered Distributed Storage IEEE Trans. Information Theory · 2019
- Secure Clustered Distributed Storage Against Eavesdropping IEEE Trans. Information Theory · 2019
- On Reusing Pilots Across Interfering Cells in Massive MIMO IEEE Trans. Wireless Comm. · 2017
- Pilot Reuse Strategy Maximizing the Weighted-Sum-Rate in Massive MIMO Systems IEEE JSAC · 2017
Funding
- · IITP — Next-Generation AI development (2025–2029)
- · NRF — Basic Research Laboratory (2024–2027)
- · NRF — Outstanding Young Scientist Award (2024–2027)