Predictive Processing and its Role in Explaining Cognition

Sunday 30/6: The Hebrew University of Jerusalem

(The Federmann Centre for the Study of Rationality / Feldman Building)

10:00-11:00 / Jakob Hohwy

11:00-12:00 / Lucia Melloni

12:00-12:30 / Coffee Break

12:30-13:30 / Regina Fabry

13:30-15:00 / Lunch Break

15:00-16:00 / Tomer Fekete

16:00-17:00 / Moshe Bar

18:30 (or 19:00) Dinner (by invitation only)

Monday 01/07: Ben-Gurion University of the Negev

(Ilse Katz Institute for Nanoscale Science and Technology / Building 51)

10:30-11:30 / Karl Friston

11:30-12:30 / Michael Gilead

12:30-14:00 / Lunch Break

14:00-15:00 / Matteo Colombo

15:00-16:00 / Merav Ahissar

16:00-16:30 / Coffee Break

16:30-17:15 / Panel or Open Discussion

18:30/18:45 / Dinner (by invitation only)

Abstracts of Talks

Show All

Hide All

Karl Friston, University College London - Me and my Markov Blanket

Karl Friston (University College London, England)

Title: Me and my Markov Blanket

Abstract:

This presentation offers a heuristic proof (and simulations of a primordial soup) suggesting that life—or biological self-organization—is an inevitable and emergent property of any (weakly mixing) random dynamical system that possesses a Markov blanket. This conclusion is based on the following arguments: if a system can be differentiated from its external milieu, heat bath or environment, then the system's internal and external states must be conditionally independent. These independencies induce a Markov blanket that separates internal and external states. This separation means that internal states will appear to minimize a free energy functional of blanket states – via a variational principle of stationary action. Crucially, this equips internal states with an information geometry, pertaining to probabilistic beliefs about something; namely external states. Interestingly, this free energy is the same quantity that is optimized in Bayesian inference and machine learning (where it is known as an evidence lower bound). In short, internal states (and their Markov blanket) will appear to model—and act on—their world to preserve their functional and structural integrity. This leads to a Bayesian mechanics, which can be neatly summarised as self-evidencing.

Jakob Hohwy, Monash University - Predictive Processing, the Free Energy Principle and Cognitive Science

Jakob Hohwy (Monash University, Melbourne, Australia)

Title: Predictive Processing, the Free Energy Principle and Cognitive Science

Abstract:

Predictive processing has emerged as a powerful paradigm in cognitive science. I review how predictive processing has led to philosophically relevant advances in our understanding of cognitive science. Behind predictive processing stands the free energy principle but its role for cognitive science is less well understood. I discuss why the free energy principle is important for cognitive science.

Matteo Colombo, Tilburg University - More Accurate but Faulty Perceptions

Matteo Colombo (Tilburg University, Netherlands)

Title: More Accurate but Faulty Perceptions

Abstract:

Patients suffering from schizophrenia are less susceptible to various perceptual illusions (and to some hallucinations too) than most healthy individuals. Yet, the processes underlying schizophrenia patients' perception have been characterised as “faulty", “false" inferences. While this type of characterisation is consistent with the idea that perceptions and perceptual processes can be evaluated as rational or irrational, it remains unclear what norms of rationality might ground the appraisal of perception—especially in the case of psychiatric patients' more accurate, but "faulty" perception. In this talk, I will consider various norms of rationality along with their supporting arguments, and I will ask in what ways-if any-these norms can illuminate the nature of the “faultiness" of schizophrenia patients' more accurate perceptual inferences.

Regina Fabry, Ruhr-Universität Bochum - What can Predictive Processing Tell Us about Spontaneous Cognition?

Regina Fabry (Ruhr-Universität Bochum, Germany)

Title: What can predictive processing tell us about spontaneous cognition?

Abstract:

For a long time, research in the cognitive sciences assumed that cognitive processes are usually task-related, goal-directed, and strictly stimulus-dependent during our waking lives. Driven by the neuroscientific discovery of the default mode network 20 years ago, recent research has explored the neuro-functional and phenomenological properties of mind-wandering and creative cognition, which are identified as distinct types of spontaneous cognition. Understanding spontaneous cognition, it is widely assumed, requires to account for the attention-driven coordination dynamics of human organisms and their local environment in a fine-grained manner. Predictive processing, with its integration of perception, action, cognition, mental simulation, and attention, provides us with valuable conceptual tools to explore the dynamical unfolding of mind-wandering and creative cognition. From this perspective, mind-wandering and creative cognition can be described in terms of distinct, temporally extended processes of precision-modulated prediction error minimisation: mind-wandering is characterised by the prioritisation of perceptual inference. In processes of creative cognition, by contrast, epistemic active inference appears to play a predominant role in prediction error minimisation. In both cases, the causal balance of perceptual inference and active inference depends on the direction of attention towards or away from the local environment, which is associated with precision estimation. The upshot is that predictive processing can help us accommodate and integrate theoretical and empirical insights into spontaneous cognition. In turn, this helps us refine our understanding of the dynamicity and flexibility of the human mind.

Merav Ahissar, Hebrew University of Jerusalem - Perceptual Priors Reveal Slow Updating in Autism and Fast Forgetting in Dyslexia

Merav Ahissar (Hebrew University of Jerusalem)

Title: Perceptual Priors Reveal Slow Updating in Autism and Fast Forgetting in Dyslexia

Abstract:

I'd like to present the relevance of the dynamics of inference for acquiring different skills--specifically, quick updating of predictions (sensory and motor plans)--within .5-~3 secs. Based on recent information, it is crucial for adequate sensory-motor actions (like driving) and for social skills (on-line interactions with people). By contrast, inferences based on long-term stimuli statistics allow the formation of complex categories allowing fluent use of language, and acquisition of expertise. We find that individuals with autism have difficulties with the former, whereas individuals with dyslexia have difficulties with the latter (Lieder et al., Nature Neurosci, 2019).

Moshe Bar, Bar-Ilan University - What’s Real? The Problem with Predictive Coding

Moshe Bar (Bar-Ilan University)

Title: What’s Real? The Problem with Predictive Coding

Abstract:

That the brain is proactive in its constant effort to generate predictions has become a mainstream idea in recent years. Indeed, predictive coding is considered by many as the best unifying framework for the operation of the brain. But our work is far from done. In my talk I will highlight a few domains in which predictive coding in general and error minimization in particular do not quite explain human cognition. These will include the influence of context, states of mind such as exploration and exploitation, first impressions, beliefs, expectations, memory distortion and attention. A common denominator for all these lies in our inherent inability to see reality as it is, “the thing-in-itself" as Kant referred to it. Directions for a potential remedy will be proposed, which might call into question whether perception as inference should indeed be the guiding principle.

Michael Gilead, Ben-Gurion University of the Negev - Above and Beyond the Concrete: The Diverse Representational Substrates of the Predictive Brain

Michael Gilead (Ben-Gurion University of the Negev)

Title: Above and Beyond the Concrete: The Diverse Representational Substrates of the Predictive Brain

Abstract:

Decades of research in philosophy and the cognitive sciences have generated rich and intricate theories of the various representational entities that subserve abstract thought. In recent years, these theories have largely given way to parsimonious models that highlight the predictive nature of cognition—the predictive brain framework. We argue that theories of the predictive brain have adopted overly simple models of mental representation, and that the lack of integration between cognitive science's rich past and its present hinders future development. We explicate the case according to which higher-order cognition relies upon a varied set of representational entities, and integrate this diversity under a unified framework that could be tested, falsified and refined through neuroscientific experimentation.

Tomer Fekete, Ben-Gurion University of the Negev - What Role will Predictive Processing Ultimately Play in Cognitive Science?

Tomer Fekete (Ben-Gurion University of the Negev)

Title: What Role will Predictive Processing Ultimately Play in Cognitive Science?

Abstract:

The framework of predictive processing (PP) has gained traction in recent years, having been suggested as an organizing principle aligning perception with action and even emotions, and some have gone as far as claiming it to establish basic awareness. In light of these developments, it could be asked what role PP can be expected to ultimately play in the cognitive sciences. We will examine three possible answers starting with the most ambitious one:

1) PP is a Theory of Consciousness.

2) PP is a Theory of Cognition.

3) PP is a plausible, perhaps, unavoidable algorithm underlying decision-making processes under uncertainty.

Considerations from the posterior explanatory power of PP suggest that it is not endowed with the requisite conceptual apparatus necessary to account for the emergence of cognition, namely a mechanism for generating new ideas and knowledge.

How is a new idea, hypothesis, or category added to an existing repertoire of ideas/hypotheses/categories, according to PP? However, in considering the relentless impinging of environmental information on the senses and the need for timely response in the face of limited computational resources, PP seems all but unavoidable as a basic decision-making mechanism under uncertainty. Thus, even though PP may not be the ultimate theory of cognition, it will likely turn out as one of the pillars of such theory.

Lucia Melloni, Max Planck Institute for Empirical Aesthetics - Knowns and unknowns of predictive computations in the human brain

Lucia Melloni (Max Planck Institute for Empirical Aesthetics, Department of Neuroscience, Germany)

Title: Knowns and unknowns of predictive computations in the human brain

Abstract:

Predictive Coding – a novel information coding framework that rests upon predictive computations - has become increasingly popular in recent years, partly because it aims to explain brain function as a whole on the basis of a small number of coding principles. Direct electrophysiological evidence for Predictive Coding computations in the human brain is scarce, even more so for the putative “canonical microcircuit" implementing Predictive Coding. Moreover, while predictions play a crucial role in Predictive Coding, it is unknown how predictions about different stimulus dimensions e.g., what an object is, where it is, and when it is presented affect perception and through which mechanisms. In this talk, I will review studies in which we have used a unique combination of functional magnetic resonance imaging, invasive electrocorticographic and intralaminar recordings, as well as lesion studies and modelling to understand how predictions are implemented and tested in the human brain.

Workshop Programme

Sunday 30/6: The Hebrew University of Jer​usalem

Monday 01/07: Ben-Gurion Univer​sity of ​​the Ne​gev

Abstracts of Talks

Karl Friston, University College London - Me and my Markov Blanket

​Jakob Hohwy, Monash University - Predictive Processing, the Free Energy Principle and Cognitive Science

​Matteo Colombo, Tilburg University - More Accurate but Faulty Perceptions

Regina Fabry, Ruhr-Universität Bochum - What can Predictive Processing Tell Us about Spontaneous Cognition?

Merav Ahissar, Hebrew University of Jerusalem - Perceptual Priors Reveal Slow Updating in Autism and Fast Forgetting in Dyslexia

Moshe Bar, Bar-Ilan University - What’s Real? The Problem with Predictive Coding

Michael Gilead, Ben-Gurion University of the Negev - Above and Beyond the Concrete: The Diverse Representational Substrates of the Predictive Brain

Tomer Fekete, Ben-Gurion University of the Negev - What Role will Predictive Processing Ultimately Play in Cognitive Science?

Lucia Melloni, Max Planck Institute for Empirical Aesthetics - Knowns and unknowns of predictive computations in the human brain

Sunday 30/6: The Hebrew University of Jerusalem

Monday 01/07: Ben-Gurion University of the Negev

Jakob Hohwy, Monash University - Predictive Processing, the Free Energy Principle and Cognitive Science

Matteo Colombo, Tilburg University - More Accurate but Faulty Perceptions