This research introduces a novel framework for verifiable commonsense reasoning in Large Language Models when answering questions using Knowledge Graphs.

• Develops a specialized approach for commonsense questions beyond factual queries
• Implements traceable reasoning procedures to verify LLM responses
• Significantly reduces hallucination by 76% compared to existing methods
• Creates the first benchmark dataset for commonsense knowledge graph QA

For security applications, this advancement offers crucial transparency in AI decision-making, enabling verification of reasoning paths and increasing trust in LLM outputs in critical information systems.

Right for Right Reasons: Large Language Models for Verifiable Commonsense Knowledge Graph Question Answering