Belief Revision: The Adaptability of Large Language Models Reasoning

TL;DR

This paper introduces Belief-R, a dataset to evaluate large language models' ability to revise beliefs with new evidence, revealing current limitations and trade-offs in their reasoning adaptability.

Contribution

The paper presents Belief-R and the delta reasoning framework, providing a new benchmark for assessing and understanding LMs' belief revision capabilities.

Findings

Most LMs struggle with belief revision tasks.

Models that update well often underperform without updates.

Significant trade-offs exist between updating and maintaining beliefs.

Abstract

The capability to reason from text is crucial for real-world NLP applications. Real-world scenarios often involve incomplete or evolving data. In response, individuals update their beliefs and understandings accordingly. However, most existing evaluations assume that language models (LMs) operate with consistent information. We introduce Belief-R, a new dataset designed to test LMs' belief revision ability when presented with new evidence. Inspired by how humans suppress prior inferences, this task assesses LMs within the newly proposed delta reasoning ($\Delta R$) framework. Belief-R features sequences of premises designed to simulate scenarios where additional information could necessitate prior conclusions drawn by LMs. We evaluate $\sim$30 LMs across diverse prompting strategies and found that LMs generally struggle to appropriately revise their beliefs in response to new information. Further, models adept at updating often underperformed in scenarios without necessary updates, highlighting a critical trade-off. These insights underscore the importance of improving LMs' adaptiveness to changing information, a step toward more reliable AI systems.