计算认知架构、双过程理论和大语言模型
PDF下载 (191)SUN Ron.计算认知架构、双过程理论和大语言模型[J].宁波大学学报(理工版),2025,38(2):17-28.DOI:10.20098/j.cnki.1001-5132.2024.1014
SUN Ron.Computational cognitive architectures, dual-process theories, and LLMs[J].Journal of Ningbo University(Natural Science & Engineering Edition),2025,38(2):17-28.DOI:10.20098/j.cnki.1001-5132.2024.1014
| Title: | Computational cognitive architectures, dual-process theories, and LLMs |
| 作者: | SUN Ron |
| Author(s): | SUN Ron |
| 关键词: | 认知架构; 大语言模型; 双过程; 心理学; 人工智能 |
| Keywords: | cognitive architecture; LLM; dual process; psychology; AI |
| 分类号: | TP18 |
| DOI: | 10.20098/j.cnki.1001-5132.2024.1014 |
| 文献标识码: | A |
| 摘要: | 介绍了一个认知科学和计算心理学领域里用于人类思维综合计算理论的计算认知架构. 这一名为Clarion的认知架构得到了心理学数据、发现和基本理论的验证. 它的一个重要理论背景是双过程理论, 这也是该模型以混合神经符号方式采用整体两级结构的原因. 此外, 鉴于人工智能和计算技术的最新进展, LLM被纳入模型中, 以更好捕捉人类的直觉和本能(以及一般意义上的隐性过程), 并进一步增强Clarion. 整合Clarion和LLM还可以帮助开发更强大、更可靠、更像人类的人工智能系统. 总的来说, 倡导采用多学科方法来开发更好的认知科学和人工智能模型. |
| Abstract: | This paper introduces a computational cognitive architecture that serves as a comprehensive computational theory of the human mind, from cognitive science and computational psychology. The cognitive architecture (named Clarion) has been justified by, and validated against, psychological data, findings, and theoretical constructs. One important theoretical background for it is the dual-process theories, which led to its overall two-level structuring in a hybrid neuro-symbolic way. Furthermore, given the recent advances in AI and computing technology, LLMs are being incorporated into the model to better capture human intuition and instinct (and implicit processes in general), in order to further enhance Clarion. Integrating Clarion and LLMs can also help to develop AI systems that are more capable, more reliable, and more human-like. Overall, the paper advocates a multidisciplinary approach towards developing better models for cognitive science and for AI. |
| 参考文献 /References: | [1] SUN R. The Cambridge handbook of computational psychology[M]. New York: Cambridge University Press, 2008. [2] SUN R. Duality of the mind[M]. Mahwah, NJ: Erlbaum, 2002. [3] SUN R. Anatomy of the mind[M]. Oxford, UK: Oxford University Press, 2016. [4] SUN R. Integrating rules and connectionism for robust commonsense reasoning[M]. New York: Wiley, 1994. [5] SUN R. Dual-process theories, cognitive architectures, and hybrid neural-symbolic models[EB/OL]. [2024-03- 18]. https://content.iospress.com/articles/neurosymbolic- artificial-intelligence/nai240720. [6] SUN R. Roles of LLMs in the overall mental architecture [EB/OL]. [2024-10-26]. https://arxiv.org/abs/2410.20037. [7] Evans J, Frankish K. In two minds: dual processes and beyond[M]. Oxford, UK: Oxford University Press, 2009. [8] REBER A S. Implicit learning and tacit knowledge[J]. Journal of Experimental Psychology: General, 1989, 118(3):219-235. [9] Kahneman D. Thinking, fast and slow[M]. New York: Farrar, Straus and Giroux, 2011. [10] HÉLIE S, SUN R. Incubation, insight, and creative problem solving: a unified theory and a connectionist model[J]. Psychological Review, 2010, 117(3):994-1024. [11] SUN R, SLUSARZ P, TERRY C. The interaction of the explicit and the implicit in skill learning: a dual-process approach[J]. Psychological Review, 2005, 112(1):159-192. [12] Pew R W, Mavor A S. Modeling human and organizational behavior: application to military simulations [M]. Washington, DC: National Academy Press, 1998. [13] Baddeley A. Working memory[M]. New York: Oxford University Press, 1986. [14] SCHACTER D L. Implicit memory: history and current status[J]. Journal of Experimental Psychology: Learning, Memory, and Cognition, 1987, 13(3):501-518. [15] Tulving E. Elements of episodic memory[M]. Oxford, UK: Clarendon Press, 1983. [16] MURRAY H A. Explorations in personality[M]. New York: Oxford University Press, 1938. [17] REISS S. Multifaceted nature of intrinsic motivation: the theory of 16 basic desires[J]. Review of General Psychology, 2004, 8(3):179-193. [18] SUN R, WILSON N. Roles of implicit processes: instinct, intuition, and personality[J]. Mind & Society, 2014, 13(1):109-134. [19] REDER L. Implicit memory and metacognition[M]. Mahwah, NJ: Erlbaum, 1996. [20] ROMERO O J, ZIMMERMAN J, STEINFELD A, et al. Synergistic integration of large language models and cognitive architectures for robust AI: an exploratory analysis[EB/OL]. [2024-08-26]. http://arxiv.org/abs/2308. 09830. [21] COSMIDES L, TOOBY J. Beyond intuition and instinct blindness: toward an evolutionarily rigorous cognitive science[J]. Cognition, 1994, 50(1/2/3):41-77. [22] DREYFUS H L, DREYFUS S E. Mind over machine: the power of human intuition and expertise in the era of the computer[M]. Oxford, UK: Blackwell, 1986. [23] SUN R. Interpreting psychological notions: a dual- process computational theory[J]. Journal of Applied Research in Memory and Cognition, 2015, 4(3):191-196. [24] MCFARLAND D. Problems of animal behaviour[M]. Singapore: Longman, 1989. [25] HASHER L, ZACKS R T. Automatic and effortful processes in memory[J]. Journal of Experimental Psychology: General, 1979, 108(3):356-388. [26] Durt C, Froese T, Fuchs T. Against AI understanding and sentience[EB/OL]. [2024-08-17]. http://philsci-archive.pitt.edu/21983/. [27] MOLLO D C, MILLIÈRE R. The vector grounding problem[EB/OL]. [2024-08-15]. http://arxiv.org/abs/2304. 01481. [28] DASGUPTA I, LAMPINEN A K, CHAN S C Y, et al. Language models show human-like content effects on reasoning tasks[EB/OL]. [2024-09-26]. http://arxiv.org/ abs/2207.07051. [29] SAPAROV A, HE H. Language models are greedy reasoners: a systematic formal analysis of chain-of- thought[EB/OL]. [2024-06-08]. http://arxiv.org/abs/2210. 01240. [30] TROTT S, JONES C, CHANG T, et al. Do large language models know what humans know?[J]. Cognitive Science, 2023, 47(7):e13309. [31] PAVLICK E. Symbols and grounding in large language models[J]. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2023, 381:20220041. [32] SØGAARD A. Understanding models understanding language[J]. Synthese, 2022, 200(6):443. [33] MACCHI L, BAGASSI M, VIALE R. Cognitive unconscious and human rationality[M]. Cambridge, MA: MIT Press, 2016. [34] SUMERS T R, YAO S Y, NARASIMHAN K, et al. Cognitive architectures for language agents[EB/OL]. [2024-07-13]. http://arxiv.org/abs/2309.02427. [35] VYGOTSKY L S. Thought and language[M]. Cambridge, MA: MIT Press, 1962. [36] XIE Y Z, XIE T, LIN M X, et al. OlaGPT: empowering LLMs with human-like problem-solving abilities[EB/OL]. [2024-07-21]. http://arxiv.org/abs/2305.16334. [37] SUN R, SESSIONS C. Learning plans without a priori knowledge[J]. Adaptive Behavior, 2000, 8(3/4):225-253. [38] PARK J S, O’BRIEN J C, CAI C J, et al. Generative agents: interactive simulacra of human behavior[EB/OL]. [2024-09-03]. http://arxiv.org/abs/2304.03442. [39] BRETZ S, SUN R. Two models of moral judgment[J]. Cognitive Science, 2018, 42(Suppl. 1):4-37. [40] SUN R, BUGROV S, DAI D. A unified framework for interpreting a range of motivation-performance phenomena [J]. Cognitive Systems Research, 2022, 71:24-40. [41] SUN R, WILSON N, LYNCH M. Emotion: a unified mechanistic interpretation from a cognitive architecture [J]. Cognitive Computation, 2016, 8(1):1-14. [42] ELMAN J L. Learning and development in neural networks: the importance of starting small[J]. Cognition, 1993, 48(1):71-99. |
| 备注/Memo: | Received date: 2024−10−31. JOURNAL OF NINGBO UNIVERSITY ( NSEE ): http://journallg.nbu.edu.cn/ Biography: SUN Ron, professor, specializing in: cognitive science. Email: sunr@rpi.edu 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ |
