Projects

We reveal that transformers trained autoregressively naturally encode causal structures—gradient attributions directly recover underlying causal graphs without any explicit causal objectives.

A novel reinforcement learning framework that enhances generalization to unseen environments through language-guided compositional causal components.

An LLM-powered autonomous agent that automates the entire causal analysis pipeline—from algorithm selection to report generation—making advanced causal methods accessible to researchers across all domains.

A millisecond-level phase locked neural feedback system based on OpenBCI

Learn to discover causality inside a large complex system without human prior

A robust deep learning pipeline for segmenting neuronal cells in microscopy images.