Can Large Language Models Reason and Plan?
arXiv:2403.04121v234.2160 citationsh-index: 24Ann n y Acad Sci
Originality Synthesis-oriented
AI Analysis
This addresses a fundamental limitation in AI reasoning for researchers and developers, but it is incremental as it builds on existing skepticism about LLM capabilities.
The paper investigates whether large language models (LLMs) can reason and plan, finding that they lack the capability for self-critiquing to correct errors, unlike humans.
While humans sometimes do show the capability of correcting their own erroneous guesses with self-critiquing, there seems to be no basis for that assumption in the case of LLMs.