Continuum-Interaction-Driven Intelligence: Human-Aligned Neural Architecture via Crystallized Reasoning and Fluid Generation

TL;DR

The paper tackles the problem that probabilistic AI, including large language models, exhibits hallucinations, unpredictability, and misalignment in professional domains. It proposes a dual-channel architecture that couples crystallized, procedural reasoning with fluid probabilistic generation to create interpretable, auditable AI, realized as the Neuromorphic Intelligence Architecture (NIA). Key contributions include redefining chain-of-thought as programmable crystallized intelligence, a task-driven modular network delineating generation versus verification, and the demonstration that multi-turn dialogue depth correlates with human alignment. The approach yields a trustworthy, domain-adaptive framework for human-AI collaboration with auditable reasoning trails, potentially transforming AI deployment in fields like healthcare and law while highlighting future challenges in automation, cross-domain transfer, and ethics.

Abstract

Current AI systems based on probabilistic neural networks, such as large language models (LLMs), have demonstrated remarkable generative capabilities yet face critical challenges including hallucination, unpredictability, and misalignment with human decision-making. These issues fundamentally stem from the over-reliance on randomized (probabilistic) neural networks-oversimplified models of biological neural networks-while neglecting the role of procedural reasoning (chain-of-thought) in trustworthy decision-making. Inspired by the human cognitive duality of fluid intelligence (flexible generation) and crystallized intelligence (structured knowledge), this study proposes a dual-channel intelligent architecture that integrates probabilistic generation (LLMs) with white-box procedural reasoning (chain-of-thought) to construct interpretable, continuously learnable, and human-aligned AI systems. Concretely, this work: (1) redefines chain-of-thought as a programmable crystallized intelligence carrier, enabling dynamic knowledge evolution and decision verification through multi-turn interaction frameworks; (2) introduces a task-driven modular network design that explicitly demarcates the functional boundaries between randomized generation and procedural control to address trustworthiness in vertical-domain applications; (3) demonstrates that multi-turn interaction is a necessary condition for intelligence emergence, with dialogue depth positively correlating with the system's human-alignment degree. This research not only establishes a new paradigm for trustworthy AI deployment but also provides theoretical foundations for next-generation human-AI collaborative systems.

Figures (1)

• Figure 1: Crystallized Reasoning.