Judea Pearl

Researchers are applying causal mediation analysis and interventions to reverse-engineer large language models into debuggable circuits, enabling teams to localize why a model fails and sometimes edit…

A new study reveals that reinforcement learning agents using causal models can perform worse than model-free agents if their causal graphs are imperfect, because learning causality from RL interaction…

Aether AI, founded by UC San Diego professor Biwei Huang, announced a $20 million seed round led by MPCi to develop Causal World Models for robotics and physical AI, aiming to replace statistical corr…