BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Bordeaux Neurocampus - ECPv4.9.10//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Bordeaux Neurocampus
X-ORIGINAL-URL:https://www.bordeaux-neurocampus.fr/en/
X-WR-CALDESC:Events for Bordeaux Neurocampus
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20240331T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20241027T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20240208T143000
DTEND;TZID=Europe/Paris:20240208T143000
DTSTAMP:20260410T121821
CREATED:20240129T134405Z
LAST-MODIFIED:20240130T074909Z
UID:167379-1707402600-1707402600@www.bordeaux-neurocampus.fr
SUMMARY:Mini-symposium - Frontiers in Bio-Inspired Learning and Intelligent Systems
DESCRIPTION:Venue: Salle de Conférence du CGFB \n\nOrganized in the frame of Hugo Chateau-Laurent’s thesis defense (February 8th\, 16h30- Broca Center) \nWith (see details below) \n\nRufin VanRullen\nMehdi Khamassi\n\nMehdi Khamassi\nTitle : Model-based and model-free reinforcement learning mechanisms in brains and robot\n\nAbstract : The reinforcement learning (RL) theory constitutes a framework for an artificial agent to learn actions that maximize rewards in the environment. It has been successfully applied to Neuroscience to account for animal neural and behavioral processes in simple laboratory tasks\, such as Pavlovian and instrumental conditioning\, and single-step economic decision-making tasks. It moreover became very popular due to its account for dopamine reward prediction error signals. However\, more complex multi-step tasks\, such as navigation and social interaction tasks\, illustrate their computational limitations.\nIn parallel\, researches in engineering\, in robotics in particular\, have emphasized the complementarity between different learning strategies when facing complex tasks\, and explored solutions to combine these different strategies. One central distinction is between model-based and model-free reinforcement learning strategies: In the former case\, an agent learns a statistical model of the effects of its actions in the environment\, and then use this model to plan sequences of actions towards desired goals. In contrast\, model-free strategies are relevant when the environment statistics are too noisy to learn a good internal model. In this case\, RL agents can rather learn local action values and adapt reactively in each state of the environment.\nIn this presentation\, I will show a series of work where we used a coordination of model-based and model-free reinforcement learning to account for a diversity of behavioral and neural observations in humans\, non-human primates and rodents in different paradigms: Navigation\, instrumental and Pavlovian conditioning. I will moreover present recent robotics results where the same algorithm with the same parameters produces optimal performance in simple navigation and social interaction tasks\, with a drastically reduced computational cost compared to classical methods. Finally\, I will show how the patterns of mental simulation within such internal models can mimic experimentally observed reactivations of the rodent hippocampus in spatial cognition tasks\, and raise new predictions for future experiments.\nRufin VanRullen\nTitle : Brain-inspired multimodal deep learning\n\nAbstract : I will describe recent efforts to design novel deep neural network architectures drawing inspiration from cognitive neuroscience. One example is the design of multimodal (e.g. text+image) architectures following the “Global Workspace” theory of cognitive science. The independent modalities converge onto a common representation space (the global workspace)\, and the shared information is then broadcast back towards the entire system. Such an architecture provides a form of referential meaning or “grounding” to each unimodal system. In addition\, the broadcast mechanism can be trained with unsupervised cycle-consistency objectives\, which makes such systems particularly attractive compared with state-of-the-art models trained on billions of paired data samples (e.g. text+image).\n
URL:https://www.bordeaux-neurocampus.fr/en/event/mini-symposium-modelisation-cognitive-et-computationnelle/
CATEGORIES:For scientists,home-event,Symposium
END:VEVENT
END:VCALENDAR