Filters

Settings

Now showing1 - 3 of 3
  • Some of the metrics are blocked by your 
    consent settings
    Item type:Publication, Access status: with files ,
    Information integration in embodied systems
    (2024)
    Langer, Carlotta
    ;
    Ay, Nihat
    Embodied agents inevitably live in constant interaction with their environment. In this thesis we model this interaction via the sensorimotor loop and analyze the different information flows among an agent’s body, brain and environment using information theoretic methods and information geometric algorithms. In particular, we highlight the interplay between the complexity of the controller of an agent and the interaction of an agent’s body with its environment.
    Publicationtype: Doctoral Thesis
      159  337
  • Some of the metrics are blocked by your 
    consent settings
    Item type:Publication, Access status: with files ,
    Invariance properties of the natural gradient in overparametrised systems
    (Springer, 2023-06)
    van Oostrum, Jesse
    ;
    Müller, Johannes
    ;
    Ay, Nihat
    The natural gradient field is a vector field that lives on a model equipped with a distinguished Riemannian metric, e.g. the Fisher–Rao metric, and represents the direction of steepest ascent of an objective function on the model with respect to this metric. In practice, one tries to obtain the corresponding direction on the parameter space by multiplying the ordinary gradient by the inverse of the Gram matrix associated with the metric. We refer to this vector on the parameter space as the natural parameter gradient. In this paper we study when the pushforward of the natural parameter gradient is equal to the natural gradient. Furthermore we investigate the invariance properties of the natural parameter gradient. Both questions are addressed in an overparametrised setting.
    Publicationtype: Journal Article
      62  233
  • Some of the metrics are blocked by your 
    consent settings
    Item type:Publication, Access status: without files ,
    Language-conditioned reinforcement learning to solve misunderstandings with action corrections
    (2022-11-17)
    Röder, Frank
    ;
    Eppe, Manfred
    Human-to-human conversation is not just talking and listening. It is an incremental process where participants continually establish a common understanding to rule out misunderstandings. Current language understanding methods for intelligent robots do not consider this. There exist numerous approaches considering non-understandings, but they ignore the incremental process of resolving misunderstandings. In this article, we present a first formalization and experimental validation of incremental action-repair for robotic instruction-following based on reinforcement learning. To evaluate our approach, we propose a collection of benchmark environments for action correction in language-conditioned reinforcement learning, utilizing a synthetic instructor to generate language goals and their corresponding corrections. We show that a reinforcement learning agent can successfully learn to understand incremental corrections of misunderstood instructions.
    Publicationtype: Conference Paper not in Proceedings
      55