提出双系统模型，让AI像人一样系统化组合推理。
A Neuroscience-Inspired Dual-Process Model of Compositional Generalization
- 用快速直觉与慢速规则双系统模拟大脑认知结构
- 在SCAN基准上达到99%以上准确率，且任务无关通用
- 可解释的规则模块适合研究推理机制与可信AI
深度学习模型难以实现系统性组合泛化，这是人类认知的核心特征。本文提出 extsc{Mirage}，一种受神经科学启发的双过程模型，通过一个快速直觉的「系统1」（元训练Transformer）和一个缓慢规则驱动的「系统2」（模式引擎）来模拟大脑新皮层与海马-前额叶回路。该模型在随机语法流上进行通用单步分解训练，在无需任务特定调整的情况下，于SCAN基准所有划分上均达到超过99%的准确率。消融实验表明，系统的组合行为源于两系统间的架构协同，特别是显式、优先级化的规则模板与迭代优化机制。与递归/循环Transformer进展一致，Mirage保持迭代神经更新的同时，将陈述性控制外化为可解释的模式模块。本工作为组合推理如何从模块化认知架构中产生提供了具体的计算模型。
原文摘要 · Abstract (English)
Deep learning models struggle with systematic compositional generalization, a hallmark of human cognition. We propose \textsc{Mirage}, a neuro-inspired dual-process model that offers a processing account for this ability. It combines a fast, intuitive ``System~1'' (a meta-trained Transformer) with a deliberate, rule-based ``System~2'' (a Schema Engine), mirroring the brain's neocortical and hippocampal--prefrontal circuits. Trained to perform general, single-step decomposition on a stream of random grammars, Mirage achieves $>$99\% accuracy on all splits of the SCAN benchmark in a task-agnostic setting. Ablations confirm that the model's systematic behavior emerges from the architectural interplay of its two systems, particularly its use of explicit, prioritized schemas and iterative refinement. In line with recent progress on recursive/recurrent Transformer approaches, Mirage preserves an iterative neural update while externalizing declarative control into an interpretable schema module. Our work provides a concrete computational model for interpreting how compositional reasoning can arise from a modular cognitive architecture.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。