Research

Towards Reliable Machine Learning (ML) in the Wild 

• Q1) How can we estimate the population density?

  • R1) Generative Models

• Q2) How can we increase the intelligence of ML?

  • R2) Agentic AI

• Q3) How can we make people trust ML?

  • R3) Causality and Invariance

R1) Generative Models

• DGMs with Disentanglement and Controllability

• DGMs for Discriminative Tasks (e.g., classification)

• Statistical Inference for DGMs

R2) Agentic AI

• Efficient Transfer Learning for Foundation Models

• Explainability of Large Multimodal Learning Models and Mechanistic Interpretability

• Agentic AI with High Reasoning Capabilities

R3) Causality and Invariance

• Invariant Representation Learning

• Causal Discovery with ML (e.g., LLMs)

• Causal Inference