Abductive Reasoning with Probabilistic Commonsense
For researchers in neurosymbolic AI and commonsense reasoning, this work provides a method to handle subjective commonsense beliefs, improving reasoning accuracy over existing approaches.
The paper addresses the challenge of varying commonsense beliefs in abductive reasoning by introducing PACS, a probabilistic neurosymbolic algorithm that samples proofs from an LLM and formal solver to aggregate conclusions. PACS outperforms chain-of-thought and prior methods across multiple benchmarks.
Recent efforts to improve the reasoning abilities of Large Language Models (LLMs) have focused on integrating formal logic solvers within neurosymbolic frameworks. A key challenge is that formal solvers lack commonsense world knowledge, preventing them from making reasoning steps that humans find obvious. Prior methods address this by using LLMs to supply missing commonsense assumptions, but these approaches implicitly assume universal agreement on such commonsense facts. In reality, commonsense beliefs vary across individuals. We propose a probabilistic framework for abductive commonsense reasoning that explicitly models this variation, aiming to determine whether most people would judge a statement as true or false. We introduce Probabilistic Abductive CommonSense (PACS), a novel algorithm that uses an LLM and a formal solver to sample proofs as observations of individuals' distinct commonsense beliefs, and aggregates conclusions across these samples. Empirically, PACS outperforms chain-of-thought reasoning, prior neurosymbolic methods, and search-based approaches across multiple benchmarks.