Dr. Thomas Parr - Active Inference Book

Topics covered
Popular Clips
Episode Highlights
Human Behavior
Active inference provides a framework for understanding human behavior and agency. explains that our sense of agency is often an inference about our actions, suggesting that much of what we perceive as intentional behavior is a post hoc explanation 1. This perspective aligns with the notion that human beliefs and desires are not consistent, but rather confabulated narratives 2.
I am inferring, I'm behaving in this way for this reason, because I've chosen to do this, because I had this goal in mind.
---
This view challenges traditional ideas of human uniqueness, proposing that our higher-level goals might also be confabulated 2.
  Â
Model Building
Building generative models involves both creativity and constraint. discusses the process of model development, emphasizing the importance of starting with a simple structure and iteratively refining it based on new data 3. This approach mirrors human cognitive processes, where models are built through exposure and adaptation.
It's not going to be worth starting from scratch every time. You can sort of build models by saying, okay, let's start with something very simple with a sort of known structure.
---
The challenge lies in balancing the complexity of models with computational tractability, a task that requires domain expertise and strategic simplification 4.
  Â
Mind Philosophy
The philosophy of mind explores the nature of consciousness and cognitive processes. and Thomas discuss the concept of sentience, questioning whether it can be reduced to simple explanations or if it possesses a unique character 5. This debate touches on the complexity of the mind, which some argue is beyond our cognitive horizon.
The dynamics of beliefs are about what's external to us and how we want to change it.
---
The discussion also highlights the predictive nature of the brain, suggesting that our perceptions are shaped by both internal models and external constraints 6.
Related Episodes


Prof. BERT DE VRIES - ON ACTIVE INFERENCE
Answers 383 questions

Dr. Sanjeev Namjoshi - Active Inference
Answers 383 questions

DR. JEFF BECK - THE BAYESIAN BRAIN
Answers 383 questions

Jonas Hübotter (ETH) - Test Time Inference
Answers 383 questions

Prof. Chris Bishop's NEW Deep Learning Textbook!
Answers 383 questions

Understanding Deep Learning - Prof. SIMON PRINCE [STAFF FAVOURITE]
Answers 383 questions

#69 DR. THOMAS LUX - Interpolation of Sparse High-Dimensional Data
Answers 383 questions

Computation, Bayesian Model Selection, Interactive Articles
Answers 383 questions
#65 Prof. PEDRO DOMINGOS [Unplugged]
Answers 383 questions

Explainability, Reasoning, Priors and GPT-3
Answers 383 questions

#49 - Meta-Gradients in RL - Dr. Tom Zahavy (DeepMind)
Answers 383 questions

#037 - Tour De Bayesian with Connor Tann
Answers 383 questions

MULTI AGENT LEARNING - LANCELOT DA COSTA
Answers 383 questions

#033 Prof. Karl Friston - The Free Energy Principle
Answers 383 questions
