Self-supervised Hierarchical Visual Reasoning with World Model

• URL: http://arxiv.org/abs/2605.17537v2
• Date: Mon, 25 May 2026 09:18:44 GMT
• Title: Self-supervised Hierarchical Visual Reasoning with World Model
• Abstract summary: 3D open-world environments with adversarial opponents remain a core challenge for reinforcement learning.<n>We propose ResDreamer, a hierarchical world model in which each higher-level layer is trained to reconstruct the residuals of the layer below.<n>This design enables progressive abstraction of increasingly sophisticated world dynamics and fosters the emergence of richer latent representations.
• Score: 82.64295546475257
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: 3D open-world environments with adversarial opponents remain a core challenge for reinforcement learning due to their vast state spaces. Effective reasoning representations are essential in such settings. While existing self-supervised visual foresight reasoning approaches often suffer from multi-step error accumulation, many recent studies resort to injecting domain-specific knowledge for more stable guidance. Our key insight is that the photorealistic fidelity of visual reasoning representations is secondary; what truly matters is providing informative, task-relevant signals. To this end, we propose ResDreamer, a hierarchical world model in which each higher-level layer is trained to reconstruct the residuals of the layer below. This design enables progressive abstraction of increasingly sophisticated world dynamics and fosters the emergence of richer latent representations. Drawing inspiration from the "Bitter Lesson", ResDreamer trains its reasoning representations in a purely self-supervised manner. The higher-level residual representations are used to modulate lower-level predictions, allowing the world model to scale effectively with only linearly increasing cross-layer communication costs. Experiments show that ResDreamer achieves state-of-the-art sample efficiency and parameter efficiency. This scalable hierarchical visual foresight reasoning architecture paves the way for more capable online RL agents in open-ended, dynamic environments. The code is accessible at https://github.com/XuYuanFei01/ResDreamer.

Related papers

• Towards Predictive, Aligned, and Scalable Robot Learning [11.830982964522605]
  We introduce Lumo-2, a latent world-action model that generates actions by reasoning over world dynamics in latent space.<n>We study latent world modelling and modality alignment, analyzing their roles in scaling laws and out-of-distribution generalization.<n>Results show that Lumo-2 consistently outperforms strong vision-language-action (VLA) and world-action model (WAM) baselines.
  arXiv  Detail & Related papers  (2026-07-13T08:53:34Z)
• Sensorimotor World Models: Perception for Action via Inverse Dynamics [59.121736411156384]
  We introduce a sensorimotor world model trained end-to-end with inverse dynamics regularization.<n>It prevents representation collapse and induces action-aligned representations.<n>It learns compact, interpretable latent spaces and enables competitive planning performance across simple 2D and 3D control tasks.
  arXiv  Detail & Related papers  (2026-06-18T11:25:16Z)
• ManipArena: Comprehensive Real-world Evaluation of Reasoning-Oriented Generalist Robot Manipulation [66.02142169323521]
  Vision-Language-Action models and world models have recently emerged as promising paradigms for general-purpose robotic intelligence.<n>Existing benchmarks are largely simulator-centric, which provide controllability but fail to capture the reality gap caused by perception noise.<n>We introduce ManipArena, a standardized evaluation framework designed to bridge simulation and real-world execution.
  arXiv  Detail & Related papers  (2026-03-30T15:06:41Z)
• Multimodal Latent Reasoning via Hierarchical Visual Cues Injection [16.779425236020433]
  This work posits that robust reasoning should evolve within a latent space, integrating multimodal signals seamlessly.<n>We propose a novel framework that instills deliberate, "slow thinking" without depending on superficial textual rationales.<n>We show that test-time scaling is effective when incorporating vision knowledge, and that integrating hierarchical information significantly enhances the model's understanding of complex scenes.
  arXiv  Detail & Related papers  (2026-02-05T06:31:12Z)
• 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)
• Clone Deterministic 3D Worlds with Geometrically-Regularized World Models [16.494281967592745]
  World models are essential for enabling agents to think, plan, and reason effectively in complex, dynamic settings.<n>Despite rapid progress, current world models remain brittle and degrade over long horizons.<n>We propose Geometrically-Regularized World Models (GRWM), which enforces that consecutive points along a natural sensory trajectory remain close in latent representation space.
  arXiv  Detail & Related papers  (2025-10-30T17:56:43Z)
• DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge [41.030494146004806]
  We propose DreamVLA, a novel VLA framework that integrates comprehensive world knowledge forecasting to enable inverse dynamics modeling.<n>DreamVLA introduces a dynamic-region-guided world knowledge prediction, integrated with the spatial and semantic cues, which provide compact yet comprehensive representations for action planning.<n>Experiments on both real-world and simulation environments demonstrate that DreamVLA achieves 76.7% success rate on real robot tasks.
  arXiv  Detail & Related papers  (2025-07-06T16:14:29Z)
• Object-centric proto-symbolic behavioural reasoning from pixels [0.0]
  We present a brain-inspired, deep-learning architecture that learns from pixels to interpret, control, and reason about its environment.<n>Results show that the agent can learn emergent conditional behavioural reasoning.<n>The proposed architecture shows how to manipulate grounded object representations, as a key inductive bias for unsupervised learning.
  arXiv  Detail & Related papers  (2024-11-26T13:54:24Z)
• MuDreamer: Learning Predictive World Models without Reconstruction [58.0159270859475]
  We present MuDreamer, a robust reinforcement learning agent that builds upon the DreamerV3 algorithm by learning a predictive world model without the need for reconstructing input signals. Our method achieves comparable performance on the Atari100k benchmark while benefiting from faster training.
  arXiv  Detail & Related papers  (2024-05-23T22:09:01Z)
• Hieros: Hierarchical Imagination on Structured State Space Sequence World Models [4.922995343278039]
  Hieros is a hierarchical policy that learns time abstracted world representations and imagines trajectories at multiple time scales in latent space. We show that our approach outperforms the state of the art in terms of mean and median normalized human score on the Atari 100k benchmark.
  arXiv  Detail & Related papers  (2023-10-08T13:52:40Z)
• A Generic Shared Attention Mechanism for Various Backbone Neural Networks [53.36677373145012]
  Self-attention modules (SAMs) produce strongly correlated attention maps across different layers. Dense-and-Implicit Attention (DIA) shares SAMs across layers and employs a long short-term memory module. Our simple yet effective DIA can consistently enhance various network backbones.
  arXiv  Detail & Related papers  (2022-10-27T13:24:08Z)
• Dynamic Inference with Neural Interpreters [72.90231306252007]
  We present Neural Interpreters, an architecture that factorizes inference in a self-attention network as a system of modules. inputs to the model are routed through a sequence of functions in a way that is end-to-end learned. We show that Neural Interpreters perform on par with the vision transformer using fewer parameters, while being transferrable to a new task in a sample efficient manner.
  arXiv  Detail & Related papers  (2021-10-12T23:22:45Z)

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.