New working paper: Causal Reasoning and Large Language Models: Opening a New Frontier for Causality; The DoWhy library for causal inference and the PyWhy organization for causal libraries and tools is growing; ICLR 2023 paper 'Modeling the Data-Generating Process is Necessary for Out-of-Distribution Generalization' was accepted with Notable-Top 25% distinction; appeared on SuperDataScience Podcast (Episode 613).
Research Experience
Head of research at Copilot Tuning; working to broaden the use of causal methods for decision-making across many application domains; projects include work at the intersection of security and machine learning; strong interest in computational social science questions and social media analyses, especially those requiring causal understanding of phenomena in health, mental health, issues of data bias, and how new technologies affect our awareness of the world and enable new kinds of information discovery and retrieval.
Background
Partner Research Manager at Microsoft, head of research at Copilot Tuning, leading efforts related to model and fine-tuning innovations and their applications to productivity scenarios. Broader research interests span causality, machine learning, and AI’s implications for people and society.