Planning with Theory of Mind

心理学 心理理论 选择(遗传算法) 平面图(考古学) 认知心理学 认知科学 心理学理论 心理干预 认知 认识论 社会心理学 人工智能 计算机科学 哲学 考古 神经科学 精神科 历史
作者
Mark K. Ho,Rebecca Saxe,Fiery Cushman
出处
期刊:Trends in Cognitive Sciences [Elsevier]
卷期号:26 (11): 959-971 被引量:77
标识
DOI:10.1016/j.tics.2022.08.003
摘要

Theory of Mind research has traditionally emphasized its predictive function (e.g., predicting someone will be angry after being stuck in traffic). Prediction tasks have dominated decades of experimental and computational research. Theory of Mind is also used to plan interventions on other minds (e.g., choosing how to cheer someone up who has been stuck in traffic) and representations used for planning will have different requirements from those only used for prediction. Research on planning emphasizes the importance of abstract and structured causal models, like Theory of Mind. Focusing on Theory of Mind for planning can illuminate a range of socio-cognitive phenomena, such as interpersonal affect regulation, impression management, pragmatic speech, and pedagogy. Understanding Theory of Mind should begin with an analysis of the problems it solves. The traditional answer is that Theory of Mind is used for predicting others’ thoughts and actions. However, the same Theory of Mind is also used for planning to change others’ thoughts and actions. Planning requires that Theory of Mind consists of abstract structured causal representations and supports efficient search and selection from innumerable possible actions. Theory of Mind contrasts with less cognitively demanding alternatives: statistical predictive models of other people’s actions, or model-free reinforcement of actions by their effects on other people. Theory of Mind is likely used to plan novel interventions and predict their effects, for example, in pedagogy, emotion regulation, and impression management. Understanding Theory of Mind should begin with an analysis of the problems it solves. The traditional answer is that Theory of Mind is used for predicting others’ thoughts and actions. However, the same Theory of Mind is also used for planning to change others’ thoughts and actions. Planning requires that Theory of Mind consists of abstract structured causal representations and supports efficient search and selection from innumerable possible actions. Theory of Mind contrasts with less cognitively demanding alternatives: statistical predictive models of other people’s actions, or model-free reinforcement of actions by their effects on other people. Theory of Mind is likely used to plan novel interventions and predict their effects, for example, in pedagogy, emotion regulation, and impression management. a mental representation that categorizes and generalizes across instances with diverse particular features. a model of the structure of relations between variables that supports counterfactual and hypothetical reasoning (i.e., one that specifies the probability of one variable setting conditioned upon intervention on another variable). actions chosen based on learned action–reward associations. Habits do not represent, and cannot be used to predict, the outcomes of actions. learning the value of an action, in context, from repeated association of the action with subsequent reward. a mechanism for action selection that relies on using a model to simulate the future consequences of potential interventions in order to choose one that maximizes expected value. a model of statistical correlations or sequences of events that can be queried to generate an expectation for one variable or state, given an observation of another variable. mechanisms and strategies for organizing memory that facilitate rapid retrieval of context- and goal-appropriate action representations in large state spaces. a mental representation that is the composition of simpler representations and whose content derives from these constituents and how they are combined. a causal model of the mind, specifying how mental states like perceptions, beliefs, and desires combine to cause actions and feelings. top-down recombination of causal elements to construct a simplified causal model that is useful for planning in context.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
田様应助坚定的藏花采纳,获得10
刚刚
ff发布了新的文献求助10
1秒前
斯文败类应助安详的凝天采纳,获得10
1秒前
1秒前
常大美女完成签到,获得积分10
2秒前
深情安青应助自然的岱周采纳,获得10
3秒前
沉静的黎昕完成签到,获得积分10
3秒前
丘比特应助称心的乘云采纳,获得10
3秒前
5秒前
阔达凝天发布了新的文献求助10
5秒前
斯文败类应助万青云采纳,获得10
5秒前
量子星尘发布了新的文献求助10
6秒前
6秒前
壮观艳完成签到,获得积分10
8秒前
8秒前
8秒前
JZ完成签到,获得积分10
9秒前
Shoshana完成签到,获得积分10
10秒前
10秒前
甜北枳完成签到,获得积分10
11秒前
承乐发布了新的文献求助10
11秒前
珊珊发布了新的文献求助10
11秒前
简单幸福完成签到 ,获得积分0
11秒前
13秒前
无花果应助adamwang采纳,获得10
13秒前
13秒前
HCT发布了新的文献求助10
13秒前
15秒前
Kondo发布了新的文献求助10
16秒前
小鱼完成签到 ,获得积分10
17秒前
一一应助有意义采纳,获得10
17秒前
17秒前
橘猫完成签到 ,获得积分10
17秒前
17秒前
17秒前
18秒前
求求你帮帮我完成签到,获得积分10
18秒前
共享精神应助珊珊采纳,获得10
18秒前
共享精神应助禹宛白采纳,获得10
18秒前
18秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Basic And Clinical Science Course 2025-2026 3000
Encyclopedia of Agriculture and Food Systems Third Edition 2000
人脑智能与人工智能 1000
花の香りの秘密―遺伝子情報から機能性まで 800
Principles of Plasma Discharges and Materials Processing, 3rd Edition 400
Pharmacology for Chemists: Drug Discovery in Context 400
热门求助领域 (近24小时)
化学 材料科学 生物 医学 工程类 计算机科学 有机化学 物理 生物化学 纳米技术 复合材料 内科学 化学工程 人工智能 催化作用 遗传学 数学 基因 量子力学 物理化学
热门帖子
关注 科研通微信公众号,转发送积分 5608256
求助须知:如何正确求助?哪些是违规求助? 4692810
关于积分的说明 14875754
捐赠科研通 4717042
什么是DOI,文献DOI怎么找? 2544147
邀请新用户注册赠送积分活动 1509105
关于科研通互助平台的介绍 1472802