Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning

• URL: http://arxiv.org/abs/2608.27549v1
• Date: Thu, 27 Aug 2026 17:57:25 GMT
• Title: Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning
• Abstract summary: We introduce Code-as-World, a paradigm that represents physical worlds through executable world representations.<n>By expressing physical composition, dynamic evolution, and visual appearance as executable code, Code-as-World provides a compact, quantitatively grounded, and controllable abstraction of the physical world.<n>We use verified executable worlds to provide scalable physical supervision for training vision-language models on quantitative physical reasoning.
• Score: 79.29811222102721
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Physical understanding and reasoning depend on forming compact and generalizable representations of the world. While modern vision-language models can recognize and explain diverse physical events, they often lack explicit representations of the underlying mechanisms-such as object states, physical parameters, and governing dynamics-needed for reliably reasoning how the world evolves and responds to interventions. In this work, we introduce Code-as-World, a paradigm that represents physical worlds through executable world representations. By expressing physical composition, dynamic evolution, and visual appearance as executable code, Code-as-World provides a compact, quantitatively grounded, and controllable abstraction of the physical world. To construct such representations from multimodal observations, such as natural-language descriptions or real-world videos, we develop an agentic discovery loop inspired by abductive reasoning, where an agent proposes, executes, renders, verifies, and iteratively refines executable world hypotheses. As a concrete application, we use verified executable worlds to provide scalable physical supervision for training vision-language models on quantitative physical reasoning. Experiments show that Code-as-World-VL achieves state-of-the-art performance on QuantiPhy and surpasses leading proprietary models, highlighting the potential of executable world representations as a scalable foundation for physical intelligence.

Related papers

• PhiZero: A World Model Built Around Physical Language [50.268271884186845]
  PhiZero is a physical world model built around physical language.<n>It infers future world evolution as a physical-language sequence and then renders the inferred transitions into videos.<n>We show its potential for realistic and interactive world modeling, fine-grained action-conditioned simulation, and zero-shot motion transfer.
  arXiv  Detail & Related papers  (2026-07-30T17:59:46Z)
• Physical Object Understanding with a Physically Controllable World Model [11.085294403458972]
  A central challenge in visual intelligence is learning the physical structure of scenes from raw videos.<n>We introduce a new class of probabilistic world models that support estimation of the probability of any visual variable.<n>We show that our world model can manipulate objects in 3D, enabling applications such as Visual Jenga.
  arXiv  Detail & Related papers  (2026-05-30T00:10:32Z)
• Visual Generation Unlocks Human-Like Reasoning through Multimodal World Models [60.543714835980325]
  Humans construct internal world models and reason by manipulating the concepts within these models.<n>Recent advances in AI approximate such human cognitive abilities, where world models are believed to be embedded within large language models.<n>This paper presents the first principled study of when and how visual generation benefits reasoning.
  arXiv  Detail & Related papers  (2026-01-27T17:40:07Z)
• Do-Undo: Generating and Reversing Physical Actions in Vision-Language Models [57.71440995598757]
  We introduce the Do-Undo task and benchmark to address a critical gap in vision-language models.<n>Do-Undo requires models to simulate the outcome of a physical action and then accurately reverse it, reflecting true cause-and-effect in the visual world.
  arXiv  Detail & Related papers  (2025-12-15T18:03:42Z)
• WorldLens: Full-Spectrum Evaluations of Driving World Models in Real World [100.68103378427567]
  Generative world models are reshaping embodied AI, enabling agents to synthesize realistic 4D driving environments that look convincing but often fail physically or behaviorally.<n>We introduce WorldLens, a full-spectrum benchmark evaluating how well a model builds, understands, and behaves within its generated world.<n>We further construct WorldLens-26K, a large-scale dataset of human-annotated videos with numerical scores and textual rationales, and develop WorldLens-Agent.
  arXiv  Detail & Related papers  (2025-12-11T18:59:58Z)
• WoW: Towards a World omniscient World model Through Embodied Interaction [83.43543124512719]
  Authentic physical intuition of the world model must be grounded in extensive, causally rich interactions with the real world.<n>We present WoW, a generative world model trained on 2 million robot interaction trajectories.<n>We establish WoWBench, a new benchmark focused on physical consistency and causal reasoning in video.
  arXiv  Detail & Related papers  (2025-09-26T17:59:07Z)
• Four Principles for Physically Interpretable World Models [1.9573380763700712]
  latent representations in world models lack direct mapping to meaningful physical quantities and dynamics.<n>We argue for a fundamental shift from physically informed to physically interpretable world models.
  arXiv  Detail & Related papers  (2025-03-04T00:19:32Z)

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.