Researchers have proposed a new framework called bipartite graphical causal models (BGCMs) for representing complex causal relationships in systems with feedback mechanisms. Unlike existing frameworks, BGCMs can accurately model the effects of different interventions on such systems. The authors demonstrate the effectiveness of BGCMs through a case study of a physical system and show that they can be used to reason about domain invariances. BGCMs generalize existing frameworks while retaining their ability to perform graphical causal reasoning. --- Why it matters: This work matters because it addresses a fundamental limitation of current causal reasoning methods, which struggle with systems at equilibrium. By providing a new framework for modeling complex causal relationships, BGCMs can help researchers and engineers better understand and analyze real-world systems. Source: https://arxiv.org/abs/2608.19831