Fundamentals of Active Inference (Chapter 7, Session 31) July 28, 2026 @ActiveInference
Fundamentals of Active Inference (Chapter 7, Session 31) July 28, 2026  @ActiveInference
Uploaded July 2026 | Updated November 2026, 1 day ago
Book: Fundamentals of Active Inference
Principles, Algorithms, and Applications of the Free Energy Principle for Engineers, by Sanjeev V. Namjoshi https://mitpress.mit.edu/9780262050951/fundamentals-of-active-inference/

Register: http://textbook-group.activeinference.institute/
Code: github.com/ActiveInferenceInstitute/fundamentals

Playlist of Textbook Group: youtube.com/playlist?list=PLNm0u2n1IwdrpoTg-9l37YIBiDrrOzTw-

Direct link to group document: coda.io/d/_dZHp_UGhVZE/Meetings_su3ToqMp

------

CHAPTERS
0:00 Intro: Chapter 7 as the hinge of the book
2:19 Looking ahead: what Chapters 8, 9 and 10 add
4:46 Continuous vs discrete state spaces and why it matters
7:00 The colour example: categorical vs continuous
8:09 Granularity, performance and runtime
9:24 Q&A: is discreteness fundamental, or just resolution?
10:30 How the 2022 textbook separated discrete and continuous
11:39 Why smooth spaces have the stronger mathematical properties
12:51 Ordinality in discrete models
16:13 Choosing units: just-noticeable differences, integers, floats
17:21 Why the equation changes with what you actually measure
19:40 Q&A: does it matter which variable is discrete or continuous?
20:50 Many maps for one territory: portfolio meta-modelling
23:05 The value of small, whimsical models when learning
24:21 Categorical vs continuous distributions: area under the curve
25:32 The Chapter 7 topic map
26:39 Memory, anticipation and attention in one framework
27:49 Generative process vs generative model in the thermostat
29:01 From action prior to action posterior
30:13 Action selection rules
31:23 When beliefs about the consequences of action are wrong
32:32 Motor babbling vs approximately correct forward models
32:55 Q&A: is action a kind of exogenous force?
33:56 Action as counteraction to an exogenous force
34:48 Worked example: heater, air conditioner and the preference sweet spot
35:43 Equation 7.18: action and exogenous force as distinct variables
37:10 The environment's state transition function
38:37 Other topologies: multi-agent and dyadic settings
39:48 What is private to each side, and what sits on the interface
40:56 7.1 Autonomous states and exogenous variables
41:32 Figure 7.2: the Bayesian network with V
42:06 Why V also connects to Y
42:42 7.2 Unifying action and perception
43:08 Page 163: the pseudo-code loop
43:41 7.3 The forward model and Algorithm 10
43:57 Figure 7.3: the canonical partition diagram
45:04 The diagram is a choice, not a law
45:51 7.4 The Bayesian thermostat
47:11 Figures 7.5 and 7.6: clamping action for the first five steps
48:02 The key result: perception changes the model, action changes the environment
49:14 Mental action and attention as action
49:51 Example 7.2: countering a static exogenous force
50:24 Figure 7.11: action and perception as divergence minimisation
50:53 7.5 Multivariate active inference in generalized coordinates
51:13 Historical review: eye movements and optimal motor control
52:06 Reflex arcs and Algorithm 11
53:08 Q&A: speed of action, epistemic foraging and local minima
55:59 Q&A: does the agent build memory?
57:02 Parametric updates, structure learning and LLM-based memory layers
58:01 Local minima, random exploration and its limits
59:16 Generalized coordinates applied to action
1:00:07 Information gain vs acting randomly
1:02:07 Wrap-up/_dZHp_UGhVZE/Meetings_su3ToqMp

------

Active Inference Institute information:
Website: https://activeinference.institute/
X: https://x.com/InferenceActive
Discord: https://discord.activeinference.institute/
Donate: https://donate.activeinference.institute/
YouTube: youtube.com/c/ActiveInference
Active Inference Livestreams: https://video.activeinference.institute/
Fundamentals of Active Inference (Chapter 7, Session 31) July 28, 2026ActInf ModelStream 020.1: Kenric Nelson: Learning extreme models with the Coupled Free EnergyStell: “Where the Theories Stop: What Free Energy and Integrated Information Actually Buy You”Patricia Craja: “Percepia: Perceptual Phenotyping from Conversation”Fundamentals of Active Inference (Chapter 10, Session 48) September 25, 2026Hyojun Choi: “Adaptive Prosociality Through Allostatic Regulation in Social Dynamics”Daniel Dragolich: “NUMEN: Bare metal inference”ActInf GuestStream 084.1 ~ The Nature of Habits and Agential Systems, Jesse GActInf GuestStream 121.1 ~ Conscious active inference (Wiest and Puniani)Fundamentals of Active Inference (Part 1 review, Session 26) July 10, 2026Fundamentals of Active Inference (Part 1 review, Session 24) July 3, 2026Ahmad Nazzal: “A Lexical T‑Maze for Active Inference”
Active Inference Institute |

Fundamentals of Active Inference (Chapter 7, Session 31) July 28, 2026

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER