MIDGARD: Self-Consistency Using Minimum Description Length for Structured Commonsense Reasoning

arXiv:2405.05189v1 Announce Type: new
Abstract: We study the task of conducting structured reasoning as generating a reasoning graph from natural language input using large language models (LLMs). Previous approaches have explored various prompting schemes, yet they suffer from error propagation due to the autoregressive nature and single-pass-based decoding, which lack error correction capability. Additionally, relying solely on a single sample may result in the omission of true nodes and edges. To counter this, we draw inspiration from self-consistency (SC), which …

abstract arxiv autoregressive commonsense cs.ai cs.cl decoding error graph language language models large language large language models llms minimum natural natural language nature prompting propagation reasoning study type