From a free energy principle to a free energy model? DOI
Patricia Palacios

Physics of Life Reviews, Journal Year: 2022, Volume and Issue: 43, P. 96 - 97

Published: Sept. 20, 2022

Language: Английский

Understanding Predictive Processing. A Review DOI Creative Commons
Michał Piekarski

Avant, Journal Year: 2021, Volume and Issue: 12(1)

Published: Jan. 1, 2021

The purpose of this paper is to provide a systematic review the Predictive Processing framework (hereinafter PP) and identify its basic theoretical difficulties.For reason, it is, primarily, polemic-critical and, secondarily, historical.I discuss main concepts, positions research issues present within ( §1-2).Next, I Bayesian-brain thesis §3) difficulty associated with §4).In §5, compare conservative radical approach PP internalist nature generative model in context Markov blankets.The possibility linking free energy principle FEP) homeostatic predictive mechanisms discussed §6.This followed by presentation PP's difficulties solving dark room problem exploration-exploitation trade-off §7).I emphasize need integrate other models frameworks cognitive science.Thus, not only discusses PP, but also provides an assessment condition light hopes placed on many researchers.The Conclusions section further challenges epistemological significance PP.

Language: Английский

Citations

11

Is the brain an organ for free energy minimisation? DOI Creative Commons
Daniel R. Williams

Philosophical Studies, Journal Year: 2021, Volume and Issue: 179(5), P. 1693 - 1714

Published: Sept. 7, 2021

Abstract Two striking claims are advanced on behalf of the free energy principle (FEP) in cognitive science and philosophy: (i) that it identifies a condition possibility existence for self-organising systems; (ii) has important implications our understanding how brain works, defining set process theories—roughly, theories structure functions neural mechanisms—consistent with minimising imperative derives as necessary feature all systems. I argue conjunction rests fallacy equivocation. The FEP can be interpreted two ways: claim about is possible to redescribe systems (the Descriptive ), such maintain their Explanatory ). Although plausibly does identify systems, no works. would have if were true, not consider various ways responding this conclusion, explore its role importance philosophy.

Language: Английский

Citations

10

Framing the predictive mind: why we should think again about Dreyfus DOI Creative Commons
Jack Reynolds

Phenomenology and the Cognitive Sciences, Journal Year: 2024, Volume and Issue: unknown

Published: May 6, 2024

Abstract In this paper I return to Hubert Dreyfus’ old but influential critique of artificial intelligence, redirecting it towards contemporary predictive processing models the mind (PP). focus on arguments about “frame problem” for cognitive systems, and his contrasting account embodied human skills expertise. The frame problem presents as a prima facie practical work in AI robotics, also computational views general, including PP. Indeed, some issues seem more acute PP, insofar seeks unify all cognition aims do so without admitting any processes or mechanisms outside scope theory. contend, however, that there is an unresolved PP concerning whether can both explain intelligent behavior minimizing prediction error with just core formal elements toolbox, adequately comprehend (or away) apparent differences between biological prediction-based notably regard establishing relevance flexible context-switching, precisely features interest indexicality, habits/skills, abductive inference. address several philosophical versions Jakob Hohwy Andy Clark, well enactive-oriented interpretations active inference coming from broadly Fristonian perspective.

Language: Английский

Citations

1

On the Origin of Cognition DOI Creative Commons
Arto Annila

Biological Theory, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 11, 2024

Language: Английский

Citations

1

Physics of creation - Symmetry breaking, (en)active inference, and unfolding statespaces DOI Open Access
Avel GUÉNIN--CARLUT

Published: Oct. 15, 2022

Physical explanation relies on the description of natural systems through construct “statespaces”, i.e. mathematical representations space their accessible states. By construction, a statespace representation must account for all determinants target system’s evolution, and any change that it does not represent be held as causally irrelevant. Therefore, although appears necessary formal tractability, entails reification physical symmetries underlying system.This is unproblematic if we hold to Platonic belief in an ordered, legible universe, whose activity explained by well-defined “natural laws” imposing observable onto Nature. However, this idea relates poorly life sciences, where organization underlies system clearly history-ladden context-sensitive. More precisely, mind seem permeated phenomenon autopoiesis (or self-creation), which they create own constitutive symmetries.We discuss here motivations law-like explanation, its intrinsic limitation “creation” mind. We then draft derived from Free Energy Principle, rearticulated “statespace-free” theory observation grounded quantum information theory. Finally, implication treatment our role agents constructing reality.

Language: Английский

Citations

5

Dosis sola facit venenum: reconceptualising biological realism DOI
Majid Davoody Beni

Biology & Philosophy, Journal Year: 2022, Volume and Issue: 37(6)

Published: Nov. 11, 2022

Language: Английский

Citations

5

Incorporating (variational) free energy models into mechanisms: the case of predictive processing under the free energy principle DOI Creative Commons
Michał Piekarski

Synthese, Journal Year: 2023, Volume and Issue: 202(2)

Published: Aug. 10, 2023

Abstract The issue of the relationship between predictive processing (PP) and free energy principle (FEP) remains a subject debate controversy within research community. Many researchers have expressed doubts regarding actual integration PP with FEP, questioning whether FEP can truly contribute significantly to mechanistic understanding or even undermine such altogether. In this paper, I present an alternative perspective. argue that, from viewpoint constraint-based mechanisms approach, imposes important constraint, namely variational energy, on architecture proposed by PP. According high-level cognitive are integral parts extensive heterarchical networks that govern physiology behavior agents. Consequently, explanations phenomena should incorporate constraints flows as relevant components, given implemented operate long is available. Within framework, contend provides constraint for explaining at least some biological described in terms Bayesian generative models minimize prediction errors.

Language: Английский

Citations

2

A Drive to Survive: the free energy principle and the meaning of life (draft) DOI Open Access
Kathryn Nave

Published: Sept. 6, 2023

[Forthcoming with MIT Press in March 2025]The Free Energy Principle proposes to reduce the goal-directed behaviour of a living agent preservation homeostasis and then analyse cognition as process pursuing this goal. I argue that stability-based account fails capture two features distinguish intentional actions biological agents from mere movements inanimate machines: namely 1) precarious dependence their bodily structure upon very metabolic processes it supports 2) potential for developmental changes structure. It is only by rejecting presumption machine-style separation between we can uniquely status systems, virtue which they may be described autonomous agents, acting towards goal own continued survival.

Language: Английский

Citations

2

Improving pathway prediction accuracy of constraints-based metabolic network models by treating enzymes as microcompartments DOI Creative Commons
Xue Yang, Zhitao Mao, Jianfeng Huang

et al.

Synthetic and Systems Biotechnology, Journal Year: 2023, Volume and Issue: 8(4), P. 597 - 605

Published: Sept. 12, 2023

Metabolic network models have become increasingly precise and accurate as the most widespread practical digital representations of living cells. The prediction functions were significantly expanded by integrating cellular resources abiotic constraints in recent years. However, if unreasonable modeling methods adopted due to a lack consideration biological knowledge, conflicts between stoichiometric other constraints, such thermodynamic feasibility enzyme resource availability, would lead distorted predictions. In this work, we investigated anomaly EcoETM, constraints-based metabolic model, introduced idea compartmentalization into analysis process. Through rational combination reactions, avoid false pathway caused unrealistic assumption free intermediate metabolites. This allowed us correct structures l-serine l-tryptophan. A specific explains application method EcoETM-like model demonstrates its potential value correcting results structure resolving conflict different incorporating evolved roles enzymes reaction compartments. Notably, work also reveals trade-off product yield feasibility. Our is great for structural improvement models.

Language: Английский

Citations

2

Structural Realism About the Free Energy Principle, the Best of Both Worlds DOI Creative Commons
Majid Davoody Beni

Journal for General Philosophy of Science, Journal Year: 2024, Volume and Issue: unknown

Published: March 26, 2024

Abstract There are realist and antirealist interpretations of the free energy principle (FEP). This paper aims to chart out a structural interpretation FEP. To do so, it draws on Worrall’s (Dialectica 43(1–2): 99–124, 1989) proposal. The general insight is that there progress at level structure theories rather than their content. enact strategy in context FEP, this will focus characterising formal continuity between fundamental equations thermodynamics—such as Boltzmann’s equation Gibbs’s equation—on one hand, Friston’s characterisation FEP other. Lack universal consensus physical character entities feature thermodynamics, information theory notwithstanding, I argue unity mathematical regiment entropy, energy. existence such provides grounds for realism about

Language: Английский

Citations

0