no code implementations • 6 Apr 2024 • Nicolas Yax, Pierre-Yves Oudeyer, Stefano Palminteri
This paper introduces PhyloLM, a method applying phylogenetic algorithms to Large Language Models to explore their finetuning relationships, and predict their performance characteristics.
1 code implementation • 13 Mar 2024 • Jérémy Perez, Corentin Léger, Marcela Ovando-Tellez, Chris Foulon, Joan Dussauld, Pierre-Yves Oudeyer, Clément Moulin-Frier
We here present a framework for simulating cultural evolution in populations of LLMs, allowing the manipulation of variables known to be important in cultural evolution, such as network structure, personality, and the way social information is aggregated and transformed.
no code implementations • 19 Feb 2024 • Grgur Kovač, Rémy Portelas, Masataka Sawayama, Peter Ford Dominey, Pierre-Yves Oudeyer
In this paper, we present a case-study about the stability of value expression over different contexts (simulated conversations on different topics), and as measured using a standard psychology questionnaire (PVQ) and a behavioral downstream task.
no code implementations • 14 Feb 2024 • Gautier Hamon, Mayalen Etcheverry, Bert Wang-Chak Chan, Clément Moulin-Frier, Pierre-Yves Oudeyer
The research field of Artificial Life studies how life-like phenomena such as autopoiesis, agency, or self-regulation can self-organize in computer simulations.
no code implementations • 16 Jan 2024 • Benjamin Clément, Hélène Sauzéon, Didier Roy, Pierre-Yves Oudeyer
In this context, the ZPDES algorithm, based on the Learning Progress Hypothesis (LPH) and multi-armed bandit machine learning techniques, sequences exercises that maximize learning progress (LP).
no code implementations • 1 Dec 2023 • Mayalen Etcheverry, Bert Wang-Chak Chan, Clément Moulin-Frier, Pierre-Yves Oudeyer
Holmes incrementally learns a hierarchy of modular representations to characterize divergent sources of diversity and uses a goal-based intrinsically-motivated exploration as the diversity search strategy.
no code implementations • 15 Oct 2023 • Julien Pourcel, Cédric Colas, Pierre-Yves Oudeyer, Laetitia Teodorescu
We here study automated problem generation in the context of the open-ended space of python programming puzzles.
1 code implementation • 17 Jul 2023 • Mayalen Etcheverry, Michael Levin, Clément Moulin-Frier, Pierre-Yves Oudeyer
Advances in bioengineering and biomedicine demand a deep understanding of the dynamic behavior of biological systems, ranging from protein pathways to complex cellular processes.
no code implementations • 15 Jul 2023 • Grgur Kovač, Masataka Sawayama, Rémy Portelas, Cédric Colas, Peter Ford Dominey, Pierre-Yves Oudeyer
We introduce the concept of perspective controllability, which refers to a model's affordance to adopt various perspectives with differing values and personality traits.
no code implementations • 15 Jul 2023 • Grgur Kovač, Rémy Portelas, Peter Ford Dominey, Pierre-Yves Oudeyer
Developmental psychologists have long-established the importance of socio-cognitive abilities in human intelligence.
no code implementations • 21 May 2023 • Cédric Colas, Laetitia Teodorescu, Pierre-Yves Oudeyer, Xingdi Yuan, Marc-Alexandre Côté
Without relying on any hand-coded goal representations, reward functions or curriculum, we show that LMA3 agents learn to master a large diversity of skills in a task-agnostic text-based environment.
no code implementations • 17 Apr 2023 • Ziang Xiao, Xingdi Yuan, Q. Vera Liao, Rania Abdelghani, Pierre-Yves Oudeyer
In this study, we explored the use of large language models (LLMs) in supporting deductive coding, a major category of qualitative analysis where researchers use pre-determined codebooks to label the data into a fixed set of codes.
no code implementations • 10 Feb 2023 • Laetitia Teodorescu, Xingdi Yuan, Marc-Alexandre Côté, Pierre-Yves Oudeyer
We show the importance of selectivity from the social peer's feedback; that experience replay needs to over-sample examples of rare goals; and that following self-generated goal sequences where the agent's competence is intermediate leads to significant improvements in final performance.
3 code implementations • 6 Feb 2023 • Thomas Carta, Clément Romac, Thomas Wolf, Sylvain Lamprier, Olivier Sigaud, Pierre-Yves Oudeyer
Using an interactive textual environment designed to study higher-level forms of functional grounding, and a set of spatial and navigation tasks, we study several scientific questions: 1) Can LLMs boost sample efficiency for online learning of various RL tasks?
1 code implementation • 14 Dec 2022 • Erwan Plantec, Gautier Hamon, Mayalen Etcheverry, Pierre-Yves Oudeyer, Clément Moulin-Frier, Bert Wang-Chak Chan
Finally, we show that Flow Lenia enables the integration of the parameters of the CA update rules within the CA dynamics, making them dynamic and localized, allowing for multi-species simulations, with locally coherent update rules that define properties of the emerging creatures, and that can be mixed with neighbouring rules.
no code implementations • 25 Nov 2022 • Rania Abdelghani, Yen-Hsiang Wang, Xingdi Yuan, Tong Wang, Pauline Lucas, Hélène Sauzéon, Pierre-Yves Oudeyer
In this context, we propose to leverage advances in the natural language processing field (NLP) and investigate the efficiency of using a large language model (LLM) for automating the production of the pedagogical content of a curious question-asking (QA) training.
no code implementations • 3 Oct 2022 • Tristan Karch, Yoann Lemesle, Romain Laroche, Clément Moulin-Frier, Pierre-Yves Oudeyer
In this paper, we investigate whether artificial agents can develop a shared language in an ecological setting where communication relies on a sensory-motor channel.
no code implementations • 22 Sep 2022 • Xingdi Yuan, Tong Wang, Yen-Hsiang Wang, Emery Fine, Rania Abdelghani, Pauline Lucas, Hélène Sauzéon, Pierre-Yves Oudeyer
Large Language Models (LLMs) have in recent years demonstrated impressive prowess in natural language generation.
no code implementations • NAACL (Wordplay) 2022 • Laetitia Teodorescu, Eric Yuan, Marc-Alexandre Côté, Pierre-Yves Oudeyer
In this extended abstract we discuss the opportunities and challenges of studying intrinsically-motivated agents for exploration in textual environments.
1 code implementation • 20 Jun 2022 • Thomas Carta, Pierre-Yves Oudeyer, Olivier Sigaud, Sylvain Lamprier
Reinforcement learning (RL) in long horizon and sparse reward tasks is notoriously difficult and requires a lot of training steps.
no code implementations • 10 Jun 2022 • Eleni Nisioti, Mateo Mahaut, Pierre-Yves Oudeyer, Ida Momennejad, Clément Moulin-Frier
Comparing the level of innovation achieved by different social network structures across different tasks shows that, first, consistent with human findings, experience sharing within a dynamic structure achieves the highest level of innovation in tasks with a deceptive nature and large search spaces.
Cultural Vocal Bursts Intensity Prediction Reinforcement Learning (RL)
no code implementations • 2 Jun 2022 • Cédric Colas, Tristan Karch, Clément Moulin-Frier, Pierre-Yves Oudeyer
Building autonomous agents able to grow open-ended repertoires of skills across their lives is a fundamental goal of artificial intelligence (AI).
no code implementations • 12 May 2022 • Iou-Jen Liu, Xingdi Yuan, Marc-Alexandre Côté, Pierre-Yves Oudeyer, Alexander G. Schwing
In order to study how agents can be taught to query external knowledge via language, we first introduce two new environments: the grid-world-based Q-BabyAI and the text-based Q-TextWorld.
1 code implementation • 26 Jan 2022 • Yoann Lemesle, Masataka Sawayama, Guillermo Valle-Perez, Maxime Adolphe, Hélène Sauzéon, Pierre-Yves Oudeyer
This suggests that the semantic word representation in the CLIP visual processing is not shared with the image representation, although the word representation strongly dominates for word-embedded images.
1 code implementation • ICLR 2022 • Paul Barde, Tristan Karch, Derek Nowrouzezahrai, Clément Moulin-Frier, Christopher Pal, Pierre-Yves Oudeyer
ABIG results in a low-level, high-frequency, guiding communication protocol that not only enables an architect-builder pair to solve the task at hand, but that can also generalize to unseen tasks.
no code implementations • ICLR 2022 • Yoann Lemesle, Masataka Sawayama, Guillermo Valle-Perez, Maxime Adolphe, Hélène Sauzéon, Pierre-Yves Oudeyer
This suggests that the semantic word representation in the CLIP visual processing is not shared with the image representation, although the word representation strongly dominates for word-embedded images.
no code implementations • 2 Jul 2021 • Grgur Kovač, Rémy Portelas, Katja Hofmann, Pierre-Yves Oudeyer
In this paper, we argue that aiming towards human-level AI requires a broader set of key social skills: 1) language use in complex and variable social contexts; 2) beyond language, complex embodied communication in multimodal settings within constantly evolving social worlds.
1 code implementation • 28 Jun 2021 • Maxime Gasse, Damien Grasset, Guillaume Gaudron, Pierre-Yves Oudeyer
We then ask the following questions: can the online and offline experiences be safely combined for learning a causal model ?
no code implementations • 25 Jun 2021 • Guillermo Valle-Pérez, Gustav Eje Henter, Jonas Beskow, André Holzapfel, Pierre-Yves Oudeyer, Simon Alexanderson
First, we present a novel probabilistic autoregressive architecture that models the distribution over future poses with a normalizing flow conditioned on previous poses as well as music context, using a multimodal transformer encoder.
1 code implementation • NeurIPS 2021 • Tristan Karch, Laetitia Teodorescu, Katja Hofmann, Clément Moulin-Frier, Pierre-Yves Oudeyer
While there is an extended literature studying how machines can learn grounded language, the topic of how to learn spatio-temporal linguistic concepts is still largely uncharted.
no code implementations • 25 May 2021 • Olivier Sigaud, Ahmed Akakzia, Hugo Caselles-Dupré, Cédric Colas, Pierre-Yves Oudeyer, Mohamed Chetouani
In the field of Artificial Intelligence, these extremes respectively map to autonomous agents learning from their own signals and interactive learning agents fully taught by their teachers.
no code implementations • 27 Apr 2021 • Grgur Kovač, Rémy Portelas, Katja Hofmann, Pierre-Yves Oudeyer
Building embodied autonomous agents capable of participating in social interactions with humans is one of the main challenges in AI.
1 code implementation • 17 Mar 2021 • Clément Romac, Rémy Portelas, Katja Hofmann, Pierre-Yves Oudeyer
Training autonomous agents able to generalize to multiple tasks is a key target of Deep Reinforcement Learning (DRL) research.
no code implementations • 11 Mar 2021 • Manfred Eppe, Pierre-Yves Oudeyer
This paper outlines a perspective on the future of AI, discussing directions for machines models of human-like intelligence.
no code implementations • 17 Dec 2020 • Cédric Colas, Tristan Karch, Olivier Sigaud, Pierre-Yves Oudeyer
Developmental RL is concerned with the use of deep RL algorithms to tackle a developmental problem -- the $intrinsically$ $motivated$ $acquisition$ $of$ $open$-$ended$ $repertoires$ $of$ $skills$.
2 code implementations • NeurIPS 2020 • Cédric Colas, Tristan Karch, Nicolas Lair, Jean-Michel Dussoux, Clément Moulin-Frier, Peter Dominey, Pierre-Yves Oudeyer
We argue that the ability to imagine out-of-distribution goals is key to enable creative discoveries and open-ended learning.
no code implementations • 16 Nov 2020 • Rémy Portelas, Clément Romac, Katja Hofmann, Pierre-Yves Oudeyer
In such complex task spaces, it is essential to rely on some form of Automatic Curriculum Learning (ACL) to adapt the task sampling distribution to a given learning agent, instead of randomly sampling tasks, as many could end up being either trivial or unfeasible.
2 code implementations • 9 Oct 2020 • Cédric Colas, Boris Hejblum, Sébastien Rouillon, Rodolphe Thiébaut, Pierre-Yves Oudeyer, Clément Moulin-Frier, Mélanie Prague
Epidemiologists model the dynamics of epidemics in order to propose control strategies based on pharmaceutical and non-pharmaceutical interventions (contact limitation, lock down, vaccination, etc).
no code implementations • 10 Aug 2020 • Grgur Kovač, Adrien Laversanne-Finot, Pierre-Yves Oudeyer
However, a currently known limitation of agents trying to maximize the diversity of sampled goals is that they tend to get attracted to noise or more generally to parts of the environments that cannot be controlled (distractors).
1 code implementation • NeurIPS 2020 • Mayalen Etcheverry, Clement Moulin-Frier, Pierre-Yves Oudeyer
Self-organization of complex morphological patterns from local interactions is a fascinating phenomenon in many natural and artificial systems.
no code implementations • ICML Workshop LaReL 2020 • Cédric Colas, Ahmed Akakzia, Pierre-Yves Oudeyer, Mohamed Chetouani, Olivier Sigaud
In the real world, linguistic agents are also embodied agents: they perceive and act in the physical world.
no code implementations • ICML Workshop LaReL 2020 • Tristan Karch, Nicolas Lair, Cédric Colas, Jean-Michel Dussoux, Clément Moulin-Frier, Peter Ford Dominey, Pierre-Yves Oudeyer
We introduce the Playground environment and study how this form of goal imagination improves generalization and exploration over agents lacking this capacity.
no code implementations • 12 Jun 2020 • Cédric Colas, Ahmed Akakzia, Pierre-Yves Oudeyer, Mohamed Chetouani, Olivier Sigaud
In the real world, linguistic agents are also embodied agents: they perceive and act in the physical world.
1 code implementation • ICLR 2021 • Ahmed Akakzia, Cédric Colas, Pierre-Yves Oudeyer, Mohamed Chetouani, Olivier Sigaud
In a second stage (L -> G), it trains a language-conditioned goal generator to generate semantic goals that match the constraints expressed in language-based inputs.
no code implementations • 13 May 2020 • Mayalen Etcheverry, Pierre-Yves Oudeyer, Chris Reinke
A central challenge is how to learn incrementally representations in order to progressively build a map of the discovered structures and re-use it to further explore.
no code implementations • 9 Apr 2020 • Laetitia Teodorescu, Katja Hofmann, Pierre-Yves Oudeyer
Recognizing precise geometrical configurations of groups of objects is a key capability of human spatial cognition, yet little studied in the deep learning literature so far.
no code implementations • 7 Apr 2020 • Rémy Portelas, Katja Hofmann, Pierre-Yves Oudeyer
A major challenge in the Deep RL (DRL) community is to train agents able to generalize over unseen situations, which is often approached by training them on a diversity of tasks (or environments).
no code implementations • 20 Mar 2020 • Tristan Karch, Cédric Colas, Laetitia Teodorescu, Clément Moulin-Frier, Pierre-Yves Oudeyer
This paper investigates the idea of encoding object-centered representations in the design of the reward function and policy architectures of a language-guided reinforcement learning agent.
no code implementations • 10 Mar 2020 • Rémy Portelas, Cédric Colas, Lilian Weng, Katja Hofmann, Pierre-Yves Oudeyer
Automatic Curriculum Learning (ACL) has become a cornerstone of recent successes in Deep Reinforcement Learning (DRL). These methods shape the learning trajectories of agents by challenging them with tasks adapted to their capacities.
2 code implementations • 21 Feb 2020 • Cédric Colas, Tristan Karch, Nicolas Lair, Jean-Michel Dussoux, Clément Moulin-Frier, Peter Ford Dominey, Pierre-Yves Oudeyer
We argue that the ability to imagine out-of-distribution goals is key to enable creative discoveries and open-ended learning.
no code implementations • 20 Feb 2020 • Clément Moulin-Frier, Pierre-Yves Oudeyer
Computational models of emergent communication in agent populations are currently gaining interest in the machine learning community due to recent advances in Multi-Agent Reinforcement Learning (MARL).
BIG-bench Machine Learning Multi-agent Reinforcement Learning +3
1 code implementation • 16 Jan 2020 • Nicolas Lair, Clément Delgrange, David Mugisha, Jean-Michel Dussoux, Pierre-Yves Oudeyer, Peter Ford Dominey
To provide such functionalities, NL interpretation in traditional assistants should be improved: (1) The intent identification system should be able to recognize new forms of known intents, and to acquire new intents as they are expressed by the user.
no code implementations • 8 Nov 2019 • Nicolas Lair, Cédric Colas, Rémy Portelas, Jean-Michel Dussoux, Peter Ford Dominey, Pierre-Yves Oudeyer
We propose LE2 (Language Enhanced Exploration), a learning algorithm leveraging intrinsic motivations and natural language (NL) interactions with a descriptive social partner (SP).
2 code implementations • 16 Oct 2019 • Rémy Portelas, Cédric Colas, Katja Hofmann, Pierre-Yves Oudeyer
We consider the problem of how a teacher algorithm can enable an unknown Deep Reinforcement Learning (DRL) student to become good at a skill over a wide range of diverse environments.
no code implementations • ICLR 2020 • Chris Reinke, Mayalen Etcheverry, Pierre-Yves Oudeyer
Using a continuous GOL as a testbed, we show that recent intrinsically-motivated machine learning algorithms (POP-IMGEPs), initially developed for learning of inverse models in robotics, can be transposed and used in this novel application area.
no code implementations • 10 Jun 2019 • Adrien Laversanne-Finot, Alexandre Péré, Pierre-Yves Oudeyer
The automatic and efficient discovery of skills, without supervision, for long-living autonomous agents, remains a challenge of Artificial Intelligence.
2 code implementations • 15 Apr 2019 • Cédric Colas, Olivier Sigaud, Pierre-Yves Oudeyer
Consistently checking the statistical significance of experimental results is the first mandatory step towards reproducible science.
no code implementations • 25 Mar 2019 • Pierre-Yves Oudeyer, George Kachergis, William Schueller
Keywords: Early language learning, Computational and robotic models, machine learning, development, embodiment, social interaction, intrinsic motivation, self-organization, dynamical systems, complexity.
1 code implementation • 15 Oct 2018 • Cédric Colas, Pierre Fournier, Olivier Sigaud, Mohamed Chetouani, Pierre-Yves Oudeyer
In open-ended environments, autonomous learning agents must set their own goals and build their own curriculum through an intrinsically motivated exploration.
no code implementations • 23 Aug 2018 • Fabien C. Y. Benureau, Pierre-Yves Oudeyer
We consider a scenario where an agent has multiple available strategies to explore an unknown environment.
1 code implementation • 4 Jul 2018 • Adrien Laversanne-Finot, Alexandre Péré, Pierre-Yves Oudeyer
Finally, we show that the measure of learning progress, used to drive curiosity-driven exploration, can be used simultaneously to discover abstract independently controllable features of the environment.
2 code implementations • 25 Jun 2018 • Pierre Fournier, Olivier Sigaud, Mohamed Chetouani, Pierre-Yves Oudeyer
In this paper, we investigate a new form of automated curriculum learning based on adaptive selection of accuracy requirements, called accuracy-based curriculum learning.
1 code implementation • 21 Jun 2018 • Cédric Colas, Olivier Sigaud, Pierre-Yves Oudeyer
Consistently checking the statistical significance of experimental results is one of the mandatory methodological steps to address the so-called "reproducibility crisis" in deep reinforcement learning.
1 code implementation • 15 May 2018 • William Schueller, Vittorio Loreto, Pierre-Yves Oudeyer
We define here a new principled measure and a new strategy, based on the beliefs of each agent on the global state of the population.
no code implementations • 19 Apr 2018 • Sao Mai Nguyen, Pierre-Yves Oudeyer
This paper presents a technical approach to robot learning of motor skills which combines active intrinsically motivated learning with imitation learning.
no code implementations • 18 Apr 2018 • Sao Mai Nguyen, Pierre-Yves Oudeyer
It has to choose actively and hierarchically at each learning episode: 1) what to learn: which outcome is most interesting to select as a goal to focus on for goal-directed exploration; 2) how to learn: which data collection strategy to use among self-exploration, mimicry and emulation; 3) once he has decided when and what to imitate by choosing mimicry or emulation, then he has to choose who to imitate, from a set of different teachers.
1 code implementation • ICLR 2018 • Alexandre Péré, Sébastien Forestier, Olivier Sigaud, Pierre-Yves Oudeyer
Intrinsically motivated goal exploration algorithms enable machines to discover repertoires of policies that produce a diversity of effects in complex environments.
no code implementations • 28 Feb 2018 • Pierre-Yves Oudeyer
What are the mechanisms of curiosity-driven learning?
1 code implementation • ICML 2018 • Cédric Colas, Olivier Sigaud, Pierre-Yves Oudeyer
In continuous action domains, standard deep reinforcement learning algorithms like DDPG suffer from inefficient exploration when facing sparse or deceptive reward problems.
no code implementations • 14 Dec 2017 • Freek Stulp, Pierre-Yves Oudeyer
Here, we formulate and study computationally the hypothesis that such patterns can emerge spontaneously as the result of a family of stochastic optimization processes (evolution strategies with covariance-matrix adaptation), without an innate encoding of a maturational schedule.
no code implementations • 5 Dec 2017 • Pierre-Yves Oudeyer
Autonomous lifelong development and learning is a fundamental capability of humans, differentiating them from current deep learning systems.
3 code implementations • 7 Aug 2017 • Sébastien Forestier, Rémy Portelas, Yoan Mollard, Pierre-Yves Oudeyer
We present an algorithmic approach called Intrinsically Motivated Goal Exploration Processes (IMGEP) to enable similar properties of autonomous learning in machines.
no code implementations • 5 Jan 2016 • Pierre-Yves Oudeyer
This article discusses open scientific challenges for understanding development and evolution of speech forms, as a commentary to Moulin-Frier et al. (Moulin-Frier et al., 2015).
no code implementations • 20 Jan 2015 • Pierre-Yves Oudeyer
To address this challenge, building and experimenting with robots modeling the growing infant brain and body is crucial.
no code implementations • 11 Oct 2013 • Benjamin Clement, Didier Roy, Pierre-Yves Oudeyer, Manuel Lopes
We present an approach to Intelligent Tutoring Systems which adaptively personalizes sequences of learning activities to maximize skills acquired by students, taking into account the limited time and motivational resources.
1 code implementation • 21 Jan 2013 • Adrien Baranes, Pierre-Yves Oudeyer
The architecture makes the robot sample actively novel parameterized tasks in the task space, based on a measure of competence progress, each of which triggers low-level goal-directed learning of the motor policy pa- rameters that allow to solve it.
no code implementations • NeurIPS 2012 • Manuel Lopes, Tobias Lang, Marc Toussaint, Pierre-Yves Oudeyer
Formal exploration approaches in model-based reinforcement learning estimate the accuracy of the currently learned model without consideration of the empirical prediction error.
Model-based Reinforcement Learning reinforcement-learning +1