Researchers have introduced two novel classes of causal models designed for decision-making agents, termed Structural Causal Decision Models (SCDMs) and Structural Causal Decision Processes (SCDPs). These models expand upon existing frameworks by explicitly representing causal relationships and allowing decisions to be constrained by their antecedents, while also accommodating open root variables. SCDPs are particularly noted for their expressiveness, surpassing POMDPs by not assuming rational belief formation and enabling the endogenous modeling of memory and variable discounting. AI
IMPACT Introduces a new framework for modeling resource-rational agents and policy simulation in digital economies.
RANK_REASON This is a research paper published on arXiv detailing new causal models for decision-making agents.
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
