A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answering

Reasoning and question answering, as fundamental cognitive functions in humans, remain significant hurdles for artificial intelligence. While large language models (LLMs) have achieved notable success, integrating explicit memory with structured reasoning capabilities remains a persistent difficulty...

Full description

Saved in:
Bibliographic Details
Main Authors: Yao Liang, Yuwei Wang, Hongjian Fang, Feifei Zhao, Yi Zeng
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-08-01
Series:Frontiers in Artificial Intelligence
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/frai.2025.1635932/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849770796163006464
author Yao Liang
Yao Liang
Yuwei Wang
Yuwei Wang
Hongjian Fang
Hongjian Fang
Feifei Zhao
Yi Zeng
Yi Zeng
Yi Zeng
Yi Zeng
Yi Zeng
author_facet Yao Liang
Yao Liang
Yuwei Wang
Yuwei Wang
Hongjian Fang
Hongjian Fang
Feifei Zhao
Yi Zeng
Yi Zeng
Yi Zeng
Yi Zeng
Yi Zeng
author_sort Yao Liang
collection DOAJ
description Reasoning and question answering, as fundamental cognitive functions in humans, remain significant hurdles for artificial intelligence. While large language models (LLMs) have achieved notable success, integrating explicit memory with structured reasoning capabilities remains a persistent difficulty. The Differentiable Neural Computer (DNC) model, despite addressing these issues to some extent, still faces challenges such as algorithmic complexity, slow convergence, and limited robustness. Inspired by the brain's learning and memory mechanisms, this paper proposes a Memory Transformation based Differentiable Neural Computer (MT-DNC) model. The MT-DNC integrates two brain-inspired memory modules—a working memory module inspired by the cognitive system that temporarily holds and processes task-relevant information, and a long-term memory module that stores frequently accessed and enduring information—within the DNC framework, enabling the autonomous transformation of acquired experiences between these memory systems. This facilitates efficient knowledge extraction and enhances reasoning capabilities. Experimental results on the bAbI question answering task demonstrate that the proposed method outperforms existing Deep Neural Network (DNN) and DNC models, achieving faster convergence and superior performance. Ablation studies further confirm that the transformation of memory from working memory to long-term memory is critical for improving the robustness and stability of reasoning. This work offers new insights into incorporating brain-inspired memory mechanisms into dialogue and reasoning systems.
format Article
id doaj-art-e4f148d9948a404d9bbd8e9e0ee46e70
institution DOAJ
issn 2624-8212
language English
publishDate 2025-08-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Artificial Intelligence
spelling doaj-art-e4f148d9948a404d9bbd8e9e0ee46e702025-08-20T03:02:52ZengFrontiers Media S.A.Frontiers in Artificial Intelligence2624-82122025-08-01810.3389/frai.2025.16359321635932A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answeringYao Liang0Yao Liang1Yuwei Wang2Yuwei Wang3Hongjian Fang4Hongjian Fang5Feifei Zhao6Yi Zeng7Yi Zeng8Yi Zeng9Yi Zeng10Yi Zeng11Brain-inspired Cognitive Intelligence Lab, Institute of Automation, Chinese Academy of Sciences, Beijing, ChinaSchool of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, ChinaBrain-inspired Cognitive Intelligence Lab, Institute of Automation, Chinese Academy of Sciences, Beijing, ChinaCenter for Long-term Artificial Intelligence, Beijing, ChinaBrain-inspired Cognitive Intelligence Lab, Institute of Automation, Chinese Academy of Sciences, Beijing, ChinaSchool of Future Technology, University of Chinese Academy of Sciences, Beijing, ChinaBrain-inspired Cognitive Intelligence Lab, Institute of Automation, Chinese Academy of Sciences, Beijing, ChinaBrain-inspired Cognitive Intelligence Lab, Institute of Automation, Chinese Academy of Sciences, Beijing, ChinaSchool of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, ChinaCenter for Long-term Artificial Intelligence, Beijing, ChinaSchool of Future Technology, University of Chinese Academy of Sciences, Beijing, ChinaKey Laboratory of Brain Cognition and Brain-inspired Intelligence Technology, Chinese Academy of Sciences, Shanghai, ChinaReasoning and question answering, as fundamental cognitive functions in humans, remain significant hurdles for artificial intelligence. While large language models (LLMs) have achieved notable success, integrating explicit memory with structured reasoning capabilities remains a persistent difficulty. The Differentiable Neural Computer (DNC) model, despite addressing these issues to some extent, still faces challenges such as algorithmic complexity, slow convergence, and limited robustness. Inspired by the brain's learning and memory mechanisms, this paper proposes a Memory Transformation based Differentiable Neural Computer (MT-DNC) model. The MT-DNC integrates two brain-inspired memory modules—a working memory module inspired by the cognitive system that temporarily holds and processes task-relevant information, and a long-term memory module that stores frequently accessed and enduring information—within the DNC framework, enabling the autonomous transformation of acquired experiences between these memory systems. This facilitates efficient knowledge extraction and enhances reasoning capabilities. Experimental results on the bAbI question answering task demonstrate that the proposed method outperforms existing Deep Neural Network (DNN) and DNC models, achieving faster convergence and superior performance. Ablation studies further confirm that the transformation of memory from working memory to long-term memory is critical for improving the robustness and stability of reasoning. This work offers new insights into incorporating brain-inspired memory mechanisms into dialogue and reasoning systems.https://www.frontiersin.org/articles/10.3389/frai.2025.1635932/fullneural turing machinememory-augmented networksreasoning and question answeringworking/long-term memorydifferentiable neural computer
spellingShingle Yao Liang
Yao Liang
Yuwei Wang
Yuwei Wang
Hongjian Fang
Hongjian Fang
Feifei Zhao
Yi Zeng
Yi Zeng
Yi Zeng
Yi Zeng
Yi Zeng
A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answering
Frontiers in Artificial Intelligence
neural turing machine
memory-augmented networks
reasoning and question answering
working/long-term memory
differentiable neural computer
title A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answering
title_full A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answering
title_fullStr A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answering
title_full_unstemmed A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answering
title_short A brain-inspired memory transformation based differentiable neural computer for reasoning-based question answering
title_sort brain inspired memory transformation based differentiable neural computer for reasoning based question answering
topic neural turing machine
memory-augmented networks
reasoning and question answering
working/long-term memory
differentiable neural computer
url https://www.frontiersin.org/articles/10.3389/frai.2025.1635932/full
work_keys_str_mv AT yaoliang abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yaoliang abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yuweiwang abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yuweiwang abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT hongjianfang abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT hongjianfang abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT feifeizhao abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng abraininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yaoliang braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yaoliang braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yuweiwang braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yuweiwang braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT hongjianfang braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT hongjianfang braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT feifeizhao braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering
AT yizeng braininspiredmemorytransformationbaseddifferentiableneuralcomputerforreasoningbasedquestionanswering