A Hippocampal-PFC Inspired Neuro-Symbolic Architecture for Contextually Anchored Question Chain Generation

ACL ARR 2025 May Submission5554 Authors

Abstract: Generating high-quality, interconnected questions remains a significant challenge in artificial intelligence (AI), particularly in applications requiring logical coherence and social relevance. Current methods often lack a cognitive foundation to ensure meaningful question relationships, limiting their effectiveness in dynamic environments. To address this gap, we propose a novel neuroscience-inspired framework, HPN-SCA, that integrates AI with theories of prefrontal cortex function, hippocampal memory retrieval, and the dynamic interplay between the Default Mode Network (DMN) and Central Executive Network (CEN). Our methodology consists of three key steps: (1) the Prefrontal Cortex Simulator, where dual models emulate the dorsolateral prefrontal cortex (DLPFC) for logical structuring and the ventromedial prefrontal cortex (VMPFC) for social contextualization to generate preliminary questions; (2) the Hippocampus Simulator, which classifies questions into Scenario-based (retrieved from knowledge bases) or Logic-based (interactively generated) chunks, mimicking memory association mechanisms; and (3) the DMN/CEN Simulator, where difficulty-based routing refines questions through either associative (DMN) or rigorous (CEN) processing. Experiments show our HPN-SCA method outperforms baselines in coherence, diversity, and human evaluation. This work integrates AI and cognitive science, enabling applications in education and conversational AI. Future work will explore additional cognitive mechanisms.

Paper Type: Long

Research Area: NLP Applications

Research Area Keywords: Neuroscience-Inspired AI, Cognitive Architecture, Hippocampal Memory Retrieval, Question Generation,

Languages Studied: English, Simplified Chinese

Submission Number: 5554