Differences between revisions 1 and 2
Revision 1 as of 2018-07-20 11:35:02
Size: 868
Comment:
Revision 2 as of 2018-07-20 12:13:25
Size: 3956
Comment:
Deletions are marked like this. Additions are marked like this.
Line 3: Line 3:
= ANR Causal (2018--2022) = = ANR CausaL (2018--2022) =
== Cognitive​ ​architectures​ ​of​ Causal​ ​Learning ==
== Architectures​ ​cognitives​ ​de​ ​l’apprentissage​ ​causale ==

Humans have an extraordinary capacity to infer cause-effect relations. In particular, we excel in
forming ​beliefs ​about the ​causal effect of actions​. Causal learning provides the basis for rational
decision-making and allows people to engage in meaningful life and social interactions. Causal learning
is a form of goal-directed learning, defined as the capacity to rapidly learn the consequence of actions
and to select behaviours according to goals and motivational state 1​ ,2​. This ability is based on internal
models of the consequence of our behaviors​ and relies on learning rules determined by the​ contingency
between actions and outcomes​. At a first approximation, contingency Δ​P ​is operationalized as the
difference between two conditional probabilities: i) P(O|A), the probability of outcome O given action A;
ii) P(O|¬A), the probability of the outcome when the action is withheld. In everyday life, people perceive
their actions as causing a given outcome if the contingency is positive, whereas they perceive them as
preventing​ ​it​ ​if​ ​negative;​ ​when​ ​P(O|A)​ ​and​ ​P(O|¬A)​ ​are​ ​equal,​ ​people​ ​report​ ​no​ ​causal​ ​effect​​ ​.
Despite the centrality of causal learning, a clear understanding of both the internal computations and neural substrates (the so-called ​cognitive architectures​) is currently missing. ​Our project will therefore
address​ ​two​ ​key​ ​questions:

 1) What are the key ​internal representations of causal beliefs and what are the ​computational processes​​ ​that​ ​enable​ ​their​ ​formation​ ​during​ ​learning?
 2) How ​​are ​​internal​ ​representations​ ​and ​​computational​​ processes​ ​​implemented​ ​​in ​​the ​​brain? [[TagAnrCausal|CausaL]]​ ​​will​ ​address​ ​these​ ​two​ ​objectives​ ​through​ ​two​ ​dedicated​ ​research​ ​work​ ​packages​ ​(WPs).
Line 9: Line 28:
|| CNRS-Aix-Marseille Université || Brovelli || Andrea || CR CNRS || 9.0 || || Coordinateur, Responsable partenaire 1. ||
|| CNRS-Aix-Marseille Université || PERRINET || Laurent || CR CNRS || 9.0 || ||
|| CNRS-Aix-Marseille Université || Brovelli || Andrea || CR CNRS || 50% || Coordinateur, Responsable partenaire 1. ||
|| CNRS-Aix-Marseille Université || PERRINET || Laurent || CR CNRS || 25% || expert in Bayesian modelling and AI ||
|| CNRS-Aix-Marseille Université || Coulon || ​Olivier || CR CNRS || 15% || neuroimaging||
|| Groupe d'Analyse et de Théorie Economique (Lyon) || Joffily || ​Mateus || IR || 30% || AI models of causal learning ||
|| Institut des Systèmes Intelligents et de Robotique (Paris) || Khamassi ​|| Mehdi || CR CNRS || 25% || RL and meta-learning models. ||
|| ​Grenoble Institut des Neurosciences (Grenoble) || Bastin || ​Julien || MC || 25% || intracerebral electroencephalography recordings ||
|| ​University of Southern California (Los Angeles, USA) || ​ Coricelli || Giorgio || || || neuroimaging and computational modelling of learning and decision making ||
|| ​University College of London (London, UK)​ ||Lagnado || ​David || || || linking causal models with neurocomputational models ||

ANR CausaL (2018--2022)

Cognitive​ ​architectures​ ​of​ Causal​ ​Learning

Architectures​ ​cognitives​ ​de​ ​l’apprentissage​ ​causale

Humans have an extraordinary capacity to infer cause-effect relations. In particular, we excel in forming ​beliefs ​about the ​causal effect of actions​. Causal learning provides the basis for rational decision-making and allows people to engage in meaningful life and social interactions. Causal learning is a form of goal-directed learning, defined as the capacity to rapidly learn the consequence of actions and to select behaviours according to goals and motivational state 1​ ,2​. This ability is based on internal models of the consequence of our behaviors​ and relies on learning rules determined by the​ contingency between actions and outcomes​. At a first approximation, contingency Δ​P ​is operationalized as the difference between two conditional probabilities: i) P(O|A), the probability of outcome O given action A; ii) P(O|¬A), the probability of the outcome when the action is withheld. In everyday life, people perceive their actions as causing a given outcome if the contingency is positive, whereas they perceive them as preventing​ ​it​ ​if​ ​negative;​ ​when​ ​P(O|A)​ ​and​ ​P(O|¬A)​ ​are​ ​equal,​ ​people​ ​report​ ​no​ ​causal​ ​effect​​ ​. Despite the centrality of causal learning, a clear understanding of both the internal computations and neural substrates (the so-called ​cognitive architectures​) is currently missing. ​Our project will therefore address​ ​two​ ​key​ ​questions:

  • 1) What are the key ​internal representations of causal beliefs and what are the ​computational processes​​ ​that​ ​enable​ ​their​ ​formation​ ​during​ ​learning?

    2) How ​​are ​​internal​ ​representations​ ​and ​​computational​​ processes​ ​​implemented​ ​​in ​​the ​​brain? CausaL​ ​​will​ ​address​ ​these​ ​two​ ​objectives​ ​through​ ​two​ ​dedicated​ ​research​ ​work​ ​packages​ ​(WPs).

Personnes impliquées dans le projet / People involved in the project :

Etablissement / Organisation

Nom / Last name

Prénom / First name

Emploi actuel / Current position

Implication dans le projet en personne.mois** / Involvement in the project (PM)

Rôle & responsabilité dans le projet / Contribution to the project

CNRS-Aix-Marseille Université

Brovelli

Andrea

CR CNRS

50%

Coordinateur, Responsable partenaire 1.

CNRS-Aix-Marseille Université

PERRINET

Laurent

CR CNRS

25%

expert in Bayesian modelling and AI

CNRS-Aix-Marseille Université

Coulon

​Olivier

CR CNRS

15%

neuroimaging

Groupe d'Analyse et de Théorie Economique (Lyon)

Joffily

​Mateus

IR

30%

AI models of causal learning

Institut des Systèmes Intelligents et de Robotique (Paris)

Khamassi ​

Mehdi

CR CNRS

25%

RL and meta-learning models.

​Grenoble Institut des Neurosciences (Grenoble)

Bastin

​Julien

MC

25%

intracerebral electroencephalography recordings

​University of Southern California (Los Angeles, USA)

​ Coricelli

Giorgio

neuroimaging and computational modelling of learning and decision making

​University College of London (London, UK)​

Lagnado

​David

linking causal models with neurocomputational models

List of pages linked to this grant:


This work was supported by ANR project "ANR Causal" ANR-XXXX.
ANR logo


TagTag TagGrants TagYear18 TagYear19 TagYear20 TagYear21

welcome: please sign in