Abstract:Exploiting large language models (LLMs) to tackle reasoning has garneredgrowing attention. It still remains highly challenging to achieve satisfactoryresults in complex logical problems, characterized by plenty of premises withinthe prompt and requiring multi-hop reasoning. In particular, the reasoningcapabilities of LLMs are brittle to disorder and distractibility. In this work,we first examine the mechanism from the perspective of information flow andreveal that LLMs exhibit failure patterns akin to human-like cognitive biaseswhen dealing with disordered and irrelevant content in reasoning tasks.However, in contrast to LLMs, disordered and irrelevant content does notsignificantly decrease human performance, as humans have a propensity todistill the most relevant information and systematically organize theirthoughts, aiding them in responding to questions. Stem from that, we furtherpropose a novel reasoning approach named Concise and Organized Perception(COP). COP carefully analyzes the given statements to identify the mostpertinent information while eliminating redundancy efficiently. It then promptsthe LLMs in a more organized form that adapts to the model's inference process.By perceiving concise and organized context, the reasoning abilities of LLMscan be better elicited. Extensive experimental results on several popularlogical benchmarks (ProofWriter, PrOntoQA, PrOntoQA-OOD, and FOLIO) and mathbenchmark (DI-GSM) show that COP significantly outperforms previousstate-of-the-art methods.

A reasoning algorithm for pd*

Fuzzy Reasoning over RDF Data Using OWL Vocabulary

Concise and Organized Perception Facilitates Large Language Models for Deductive Reasoning.

The expressive language ALCNHR+K(D) for knowledge reasoning

Partial Semantics of Argumentation: Basic Properties and Empirical

Pellet: A practical OWL-DL reasoner

Reasoning with Large Scale Ontologies in Fuzzy pD* Using MapReduce

Finding all justifications of OWL entailments using TMS and MapReduce.

A Relevance-Directed Algorithm for Finding Justifications of DL Entailments

New reasoning algorithm based on EFALC

The Datalog(Dl) Combination Of Deduction Rules And Description Logics

Large Scale Fuzzy Pd* Reasoning Using Mapreduce

Paraconsistent Reasoning with OWL - Algorithms and the ParOWL Reasoner ?

A New Method of Finding All Justifications in OWL 2 EL

Algorithms for Paraconsistent Reasoning with OWL

A Modularization-Based Approach to Finding All Justifications for OWL DL Entailments

Commonsense Reasoning in a Deeper Way: By Discovering Relations between Predicates.

Deciding Query Entailment In Fuzzy Owl Lite Ontologies

Combining Proofs for Description Logic and Concrete Domain Reasoning (Technical Report)

Alc(P)(U): An Integration Of Description Logic And General Rules

Inference, Optimization, and Analysis of an Evidential Reasoning Rule-Based Modeling Approach