如何开始
从生态系统参考开始，启动仓库、公开直播、播客和学习小组。网站提供稳定的公共地图，而 Discord 和定期活动则提供同步参与。
此页面由机器翻译成中文。 查看英文原版。
寻求结构化入门点的学生和研究人员
学习与研究
教科书小组、视频、播客、课程、研究资源、实现和技术基础。
在本页
学习与研究 指南
学习资源引导新手和有经验的参与者向活参考、直播、Active Inference洞察、教科书小组、阅读材料、代码实现和特定领域课程。
核心阅读材料
学习路径包括托马斯·帕尔、乔瓦尼·佩祖洛和卡尔·J·弗里斯特恩的2022年教科书，《麦克斯韦·拉姆斯特德2023年摘要》、《杰雷德·图米尔的《在主动推断中上手》》、《贝伦·米利奇的论文仓库》，以及关于主动推断本体论和团队的工作。
课程和小组
学习表面包括积极推断基础、教科书小组、社会科学课程、物理学课程、RxInfer.jl 学习小组、季节学校和理论神经生物学对话。
实现
实现路径包括RxInfer.jl、PyMDP、SPM、符号认知机器人学以及开放仓库。这些支持实际建模和实验。
技术树探索者
探索「主动推断」学习图
一个交互式地图，展示了积极推断和自由能量原理技术树中的概念、方法和工具。选择一个节点以突出显示其邻居；按关系类型过滤。
{"nodes":[{"id":"active-inference:free-energy-principle","label":"Free Energy Principle","type":"concept","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:active-inference","label":"Active Inference","type":"concept","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:predictive-processing","label":"Predictive Processing","type":"concept","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:markov-blankets","label":"Markov Blankets","type":"concept","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:variational-bayes","label":"Variational Bayes","type":"method","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:generative-models","label":"Generative Models","type":"concept","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:precision-weighting","label":"Precision Weighting","type":"concept","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:expected-free-energy","label":"Expected Free Energy","type":"concept","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:policy-selection","label":"Policy Selection","type":"method","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:perception","label":"Perception as Inference","type":"application","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:action","label":"Action as Active Inference","type":"application","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:learning","label":"Learning as Model Update","type":"application","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:accessibility-value","label":"Accessibility","type":"value","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:rigor-value","label":"Scientific Rigor","type":"value","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:applicability-value","label":"Real-World Applicability","type":"value","meta":"Tree: active-inference · Maturity: established"},{"id":"active-inference:pymdp","label":"pymdp","type":"tool","meta":"Tree: active-inference · Maturity: emerging"},{"id":"active-inference:rxinfer","label":"RxInfer.jl","type":"tool","href":"../projects/rxinfer/","meta":"Tree: active-inference · Maturity: emerging"},{"id":"active-inference:karl-friston","label":"Karl Friston","type":"person","meta":"Tree: active-inference · Maturity: established"},{"id":"free-energy-principle:information-theory","label":"Information Theory","type":"research_area","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:statistical-mechanics","label":"Statistical Mechanics","type":"research_area","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:bayesian-inference","label":"Bayesian Inference","type":"method","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:entropy","label":"Entropy","type":"concept","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:surprise-minimization","label":"Surprise Minimization","type":"concept","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:fep-core","label":"Free Energy Principle (Core)","type":"concept","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:self-organization","label":"Self-Organization","type":"concept","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:autopoiesis","label":"Autopoiesis","type":"concept","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:perception-action-loop","label":"Perception-Action Loop","type":"concept","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:homeostasis","label":"Homeostasis","type":"concept","meta":"Tree: free-energy-principle · Maturity: established"},{"id":"free-energy-principle:allostasis","label":"Allostasis","type":"concept","meta":"Tree: free-energy-principle · Maturity: advanced"},{"id":"free-energy-principle:consciousness-theories","label":"Consciousness Theories","type":"research_area","meta":"Tree: free-energy-principle · Maturity: emerging"}],"edges":[{"source":"active-inference:free-energy-principle","target":"active-inference:active-inference","relation":"enables"},{"source":"active-inference:free-energy-principle","target":"active-inference:markov-blankets","relation":"part_of"},{"source":"active-inference:free-energy-principle","target":"active-inference:variational-bayes","relation":"implements"},{"source":"active-inference:free-energy-principle","target":"active-inference:generative-models","relation":"part_of"},{"source":"active-inference:predictive-processing","target":"active-inference:free-energy-principle","relation":"extends"},{"source":"active-inference:active-inference","target":"active-inference:precision-weighting","relation":"part_of"},{"source":"active-inference:active-inference","target":"active-inference:expected-free-energy","relation":"part_of"},{"source":"active-inference:expected-free-energy","target":"active-inference:policy-selection","relation":"enables"},{"source":"active-inference:active-inference","target":"active-inference:perception","relation":"applies_to"},{"source":"active-inference:active-inference","target":"active-inference:action","relation":"applies_to"},{"source":"active-inference:active-inference","target":"active-inference:learning","relation":"applies_to"},{"source":"active-inference:pymdp","target":"active-inference:active-inference","relation":"implements"},{"source":"active-inference:rxinfer","target":"active-inference:active-inference","relation":"implements"},{"source":"active-inference:karl-friston","target":"active-inference:free-energy-principle","relation":"authored_by"},{"source":"active-inference:accessibility-value","target":"active-inference:active-inference","relation":"governs"},{"source":"active-inference:rigor-value","target":"active-inference:free-energy-principle","relation":"governs"},{"source":"active-inference:applicability-value","target":"active-inference:perception","relation":"governs"},{"source":"active-inference:applicability-value","target":"active-inference:action","relation":"governs"},{"source":"active-inference:variational-bayes","target":"active-inference:generative-models","relation":"prerequisite"},{"source":"active-inference:markov-blankets","target":"active-inference:generative-models","relation":"prerequisite"},{"source":"free-energy-principle:information-theory","target":"free-energy-principle:fep-core","relation":"part_of"},{"source":"free-energy-principle:statistical-mechanics","target":"free-energy-principle:fep-core","relation":"part_of"},{"source":"free-energy-principle:bayesian-inference","target":"free-energy-principle:fep-core","relation":"implements"},{"source":"free-energy-principle:entropy","target":"free-energy-principle:surprise-minimization","relation":"prerequisite"},{"source":"free-energy-principle:surprise-minimization","target":"free-energy-principle:fep-core","relation":"instantiated_by"},{"source":"free-energy-principle:fep-core","target":"free-energy-principle:self-organization","relation":"enables"},{"source":"free-energy-principle:autopoiesis","target":"free-energy-principle:self-organization","relation":"related_to"},{"source":"free-energy-principle:fep-core","target":"free-energy-principle:homeostasis","relation":"applies_to"},{"source":"free-energy-principle:allostasis","target":"free-energy-principle:homeostasis","relation":"extends"},{"source":"free-energy-principle:perception-action-loop","target":"free-energy-principle:fep-core","relation":"part_of"},{"source":"free-energy-principle:fep-core","target":"free-energy-principle:consciousness-theories","relation":"applies_to"},{"source":"free-energy-principle:bayesian-inference","target":"free-energy-principle:entropy","relation":"prerequisite"},{"source":"free-energy-principle:information-theory","target":"free-energy-principle:entropy","relation":"prerequisite"}]}
关键表面
一目了然的学习与研究
教科书小组
通过「积极推断」基础的协作之路。
直播
公开更新、研讨会、模型直播、客座直播和教育录像。
积极推断见解
通过公共媒体渠道提供的对话式播客内容。
研究资源
参考、仓库、论文、本体论和实现链接。
学习地图
通过概念、方法和应用的结构化路径。
实现
RxInfer.jl、PyMDP、SPM以及符号认知机器人技术。
思想与方法
从学习图中提取的概念和方法
从“主动推断”和“自由能量原则”技术树节点中精心挑选的简洁集合。
| Idea | Type | Maturity | Summary | Tags | Tree |
|---|---|---|---|---|---|
| Action as Active Inference | Application | Established | Action as the process of changing the world to fulfill predictions, minimizing free energy by acting on the environment rather than updating beliefs. | action, embodied, motor | Active Inference |
| Learning as Model Update | Application | Established | Learning as the slow update of generative model parameters to minimize long-term free energy, encompassing synaptic plasticity and structure learning. | learning, model-update, plasticity | Active Inference |
| Perception as Inference | Application | Established | Perception understood as Bayesian inference — updating internal generative models to explain away prediction errors from sensory input. | inference, perception, sensory | Active Inference |
| Active Inference | Concept | Established | The process theory derived from the Free Energy Principle: agents act to fulfill predictions and minimize expected free energy through perception and action. | active-inference, process-theory | Active Inference |
| Allostasis | Concept | Advanced | Predictive homeostasis — anticipatory regulation that adjusts setpoints based on expected future demands rather than reacting to deviations. | anticipatory, predictive, regulation | Free Energy Principle |
| Autopoiesis | Concept | Established | Self-producing systems as defined by Maturana and Varela — living systems that continuously produce and maintain themselves. | maturana, self-production, varela | Free Energy Principle |
| Entropy | Concept | Established | A measure of uncertainty or disorder in a probability distribution, connecting information theory to thermodynamics. | disorder, entropy, uncertainty | Free Energy Principle |
| Expected Free Energy | Concept | Established | The objective function for policy selection in active inference, combining epistemic (information-seeking) and pragmatic (goal-seeking) drives. | efe, epistemic, policy-selection, pragmatic | Active Inference |
相关资源
此页的公共链接
外部链接从共享注册表解析，这样面向访问者的目的地保持集中和可检查。
ActiveInference.org landing page
Audience: Newcomer
Current public landing page for the Institute, including mission, vision, open-science framing, and the Get Started pathway.
officialoverviewmission
Weekly updates
Audience: Newcomer
Recurring public update surface for activity orientation and current work.
communityweekly
Start repository
Audience: Learner
Learning repository for getting started with Active Inference materials and pathways.
learningstart-repo
Ecosystem shortlink
Audience: Researcher
Public ecosystem shortlink for Institute context, projects, activities, and conceptual maps.
researchecosystem
START documentation site
Audience: Learner
Public START documentation site for tailored Active Inference learning and curriculum generation.
start-repolearningcurriculum
YouTube channel
Audience: Newcomer
Public video channel for seminars, streams, educational sessions, and community recordings.
mediayoutube
YouTube handle
Audience: Newcomer
Direct handle for the public YouTube presence.
mediayoutube-handle
Videos and podcasts shortlink
Audience: Newcomer
Public videos and podcasts shortlink for browsing recordings by format and topic.
mediavideo
Podcast feed
Audience: Newcomer
Podcast destination for Active Inference conversations and interviews.
mediapodbean
GitHub organization
Audience: Developer
Public GitHub organization for Institute repositories and open-source work.
projectsgithub-org
Generalized Notation Notation repository
Audience: Developer
Generalized Notation Notation project repository for model communication.
projectsgnn
Active Inference Journal repository
Audience: Researcher
Research and publication-oriented repository for the Active Inference Journal.
researchjournal
官方页面
官方机构表面
ActiveInference.org landing page
Audience: Newcomer
Current public Institute landing page with mission, vision, nonprofit status, community metrics, and Get Started pathway.
officialmissionoverview
Education overview
Audience: Learner
Official education entry point, currently resolving to courses.
officialeducationcourses
Courses
Audience: Learner
Official courses page for learning Active Inference.
officialcourseslearning
Textbook Group
Audience: Learner
Official Textbook Group page.
officialtextbooklearning
Livestreams
Audience: Learner
Official livestreams page.
officiallivestreamsmedia
Research overview
Audience: Researcher
Official research entry point, currently resolving to the research overview.
officialresearch
Research overview page
Audience: Researcher
Official research overview page.
officialresearchoverview
RxInfer
Audience: Developer
Official RxInfer learning page.
officialrxinferimplementation
仓库
相关开源仓库
act_inf_metaanalysis
Audience: Researcher
Computational meta-analysis of Active Inference literature with nanopublication and knowledge-graph outputs.
researchknowledgetex
ActInf_RxInfer
Audience: Learner
Active Inference and RxInfer.jl
learningimplementationunclassified
Active_Inference_Ontology
Audience: Researcher
Ontology-oriented repository for shared Active Inference concepts and decentralized science knowledge infrastructure.
researchknowledgeunclassified
ActiveInferenceCategoryTheory
Audience: Researcher
Active Inference & Category Theory
researchknowledgeunclassified
ActiveInferenceImplementations
Audience: Learner
For learning and exploring Probabilistic Inference
learningimplementationjupyter-notebook
ActiveInferenceJournal
Audience: Researcher
Public content repository for the Active Inference Journal and related publication infrastructure.
researchknowledgehtml
AEOS
Audience: Researcher
Active Entity Ontology for Science
researchknowledgeunclassified
CEREBRUM
Audience: Researcher
Case-enabled Bayesian reasoning engine for unified modeling and cognitive representations.
researchknowledgepython