Learning to Perceive the World Through Control: Empowerment-Based Representation Learning

• URL: http://arxiv.org/abs/2605.30656v1
• Date: Thu, 28 May 2026 23:28:20 GMT
• Title: Learning to Perceive the World Through Control: Empowerment-Based Representation Learning
• Abstract summary: We show that empowerment leads agents to learn an implicit, control-centric model of the world.<n>Our analysis highlights the importance of learning representations through interaction rather than from passive datasets.
• Score: 26.8155637946433
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In many practical reinforcement learning environments, observations are far higher-dimensional than the variables that matter for control. In this work, we ask: can we learn representations that capture only control-relevant features of the environment? We study this question through the empowerment objective, which maximizes an agent's influence over the environment and is widely used for unsupervised skill learning. We show that empowerment agents induce two distinct representations -- forward and backward -- that capture complementary aspects of the state, and both of which are invariant to control-irrelevant features. Thus, empowerment maximization leads agents to learn an implicit, control-centric model of the world. Our analysis highlights the importance of learning representations through interaction rather than from passive datasets: interaction aimed at maximizing control is essential for learning useful invariance properties, a perspective that aligns closely with the causal learning literature.

Related papers

• Revisiting Multi-Task Visual Representation Learning [52.93947931352643]
  We introduce MTV, a principled multi-task visual pretraining framework.<n>We leverage high-capacity "expert" models to synthesize dense, structured pseudo-labels at scale.<n>Our results demonstrate that MTV achieves "best-of-both-worlds" performance.
  arXiv  Detail & Related papers  (2026-01-20T11:59:19Z)
• Multimodal Visual-Tactile Representation Learning through Self-Supervised Contrastive Pre-Training [0.850206009406913]
  MViTac is a novel methodology that leverages contrastive learning to integrate vision and touch sensations in a self-supervised fashion. By availing both sensory inputs, MViTac leverages intra and inter-modality losses for learning representations, resulting in enhanced material property classification and more adept grasping prediction.
  arXiv  Detail & Related papers  (2024-01-22T15:11:57Z)
• Can Self-Supervised Representation Learning Methods Withstand Distribution Shifts and Corruptions? [5.706184197639971]
  Self-supervised learning in computer vision aims to leverage the inherent structure and relationships within data to learn meaningful representations. This work investigates the robustness of learned representations of self-supervised learning approaches focusing on distribution shifts and image corruptions.
  arXiv  Detail & Related papers  (2023-07-31T13:07:56Z)
• Vision-Based Manipulators Need to Also See from Their Hands [58.398637422321976]
  We study how the choice of visual perspective affects learning and generalization in the context of physical manipulation from raw sensor observations. We find that a hand-centric (eye-in-hand) perspective affords reduced observability, but it consistently improves training efficiency and out-of-distribution generalization.
  arXiv  Detail & Related papers  (2022-03-15T18:46:18Z)
• Information is Power: Intrinsic Control via Information Capture [110.3143711650806]
  We argue that a compact and general learning objective is to minimize the entropy of the agent's state visitation estimated using a latent state-space model. This objective induces an agent to both gather information about its environment, corresponding to reducing uncertainty, and to gain control over its environment, corresponding to reducing the unpredictability of future world states.
  arXiv  Detail & Related papers  (2021-12-07T18:50:42Z)
• Learning to Improve Representations by Communicating About Perspectives [0.0]
  We present aminimal architecture comprised of a population of autoencoders. We show that our proposed architectureallows the emergence of aligned representations. Results demonstrate how communication from subjective perspec-tives can lead to the acquisition of more abstract representations in multi-agent systems.
  arXiv  Detail & Related papers  (2021-09-20T09:30:13Z)
• Visual Adversarial Imitation Learning using Variational Models [60.69745540036375]
  Reward function specification remains a major impediment for learning behaviors through deep reinforcement learning. Visual demonstrations of desired behaviors often presents an easier and more natural way to teach agents. We develop a variational model-based adversarial imitation learning algorithm.
  arXiv  Detail & Related papers  (2021-07-16T00:15:18Z)
• Self-Supervised Learning Across Domains [33.86614301708017]
  We propose to apply a similar approach to the problem of object recognition across domains. Our model learns the semantic labels in a supervised fashion, and broadens its understanding of the data by learning from self-supervised signals on the same images. This secondary task helps the network to focus on object shapes, learning concepts like spatial orientation and part correlation, while acting as a regularizer for the classification task.
  arXiv  Detail & Related papers  (2020-07-24T06:19:53Z)
• Self-supervised Learning from a Multi-view Perspective [121.63655399591681]
  We show that self-supervised representations can extract task-relevant information and discard task-irrelevant information. Our theoretical framework paves the way to a larger space of self-supervised learning objective design.
  arXiv  Detail & Related papers  (2020-06-10T00:21:35Z)
• Mutual Information-based State-Control for Intrinsically Motivated Reinforcement Learning [102.05692309417047]
  In reinforcement learning, an agent learns to reach a set of goals by means of an external reward signal. In the natural world, intelligent organisms learn from internal drives, bypassing the need for external signals. We propose to formulate an intrinsic objective as the mutual information between the goal states and the controllable states.
  arXiv  Detail & Related papers  (2020-02-05T19:21:20Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.