MuSR: Testing the Limits of Chain-of-thought with Multistep Soft Reasoning

TL;DR

MuSR is a new dataset designed to evaluate large language models' multistep reasoning abilities in complex, realistic narratives, revealing current limitations of chain-of-thought prompting.

Contribution

The paper introduces MuSR, a novel neurosymbolic synthetic-to-natural dataset for testing multistep reasoning in LLMs, scalable and more realistic than existing benchmarks.

Findings

GPT-4 struggles with complex reasoning tasks in MuSR

Chain-of-thought prompting shows gaps in robustness

MuSR challenges current LLM reasoning capabilities

Abstract

While large language models (LLMs) equipped with techniques like chain-of-thought prompting have demonstrated impressive capabilities, they still fall short in their ability to reason robustly in complex settings. However, evaluating LLM reasoning is challenging because system capabilities continue to grow while benchmark datasets for tasks like logical deduction have remained static. We introduce MuSR, a dataset for evaluating language models on multistep soft reasoning tasks specified in a natural language narrative. This dataset has two crucial features. First, it is created through a novel neurosymbolic synthetic-to-natural generation algorithm, enabling the construction of complex reasoning instances that challenge GPT-4 (e.g., murder mysteries roughly 1000 words in length) and which can be scaled further as more capable LLMs are released. Second, our dataset instances are free text narratives corresponding to real-world domains of reasoning; this makes it simultaneously much more challenging than other synthetically-crafted benchmarks while remaining realistic and tractable for human annotators to solve with high accuracy. We evaluate a range of LLMs and prompting techniques on this dataset and characterize the gaps that remain for techniques like chain-of-thought to perform robust reasoning.