Essay.03-How.to.Model.Causality.From.Philosophy.to.AI-ZhuoyingLi-2100017764
Keywords: causality, causal reasoning
Abstract: This essay explores the concept of causality from philosophical and mathematical standpoints, studying its development from Hume to Lewis, and onward to the use of probability in the creation of causal models. It considers the advantages and drawbacks of Rubin's Causal Model (RCM) and Structural Causal Model (SCM). Inspired by Hume's Regularity Theory of Causation (RTC), it suggests the amalgamation of the probabilistic approach with Multimodal Large Language Models (MLLMs) as an innovative solution to accurately analyze event correlations and capture complex spatiotemporal contexts, thereby unpacking causality in a more comprehensive manner.
Submission Number: 69
Loading