mlsanityData to 7 Oct 2026

ContPhy: Continuum Physical Concept Learning and Reasoning from Videos

arXiv 2402.06119 ↗First posted 9 Feb 2024cs.CVcited by 25

On the map: LLM reasoning → Spatial and embodied reasoning

Abstract (arXiv)

We introduce the Continuum Physical Dataset (ContPhy), a novel benchmark for assessing machine physical commonsense. ContPhy complements existing physical reasoning benchmarks by encompassing the inference of diverse physical properties, such as mass and density, across various scenarios and predicting corresponding dynamics. We evaluated a range of AI models and found that they still struggle to achieve satisfactory performance on ContPhy, which shows that the current AI models still lack physical commonsense for the continuum, especially soft-bodies, and illustrates the value of the proposed dataset. We also introduce an oracle model (ContPRO) that marries the particle-based physical dynamic models with the recent large language models, which enjoy the advantages of both models, precise dynamic predictions, and interpretable reasoning. ContPhy aims to spur progress in perception and reasoning within diverse physical settings, narrowing the divide between human and machine intelligence in understanding the physical world. Project page: https://physical-reasoning-project.github.io

Key facts. Cited by 25 papers in the map (data through 7 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 6 of its 54 references that are in the map. First posted on arXiv 9 Feb 2024.

Builds on

6 of 54 references.

References · 6

  1. 2022ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understandingcited by 262
  2. 2020ThreeDWorld: A Platform for Interactive Multi-Modal Physical Simulationcited by 249
  3. 2018Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluidscited by 219
  4. 2018IntPhys: A Framework and Benchmark for Visual Intuitive Physics Reasoningcited by 72
  5. 2023Physion++: Evaluating Physical Scene Understanding that Requires Online Inference of Different Physical Propertiescited by 53
  6. 2020CRAFT: A Benchmark for Causal Reasoning About Forces and inTeractionscited by 44

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 4; 2025: 8; 2026: 13.

25 papers.

Sort:

By area:

  1. 2025QuantiPhy: A Quantitative Benchmark Evaluating Physical Reasoning Abilities of Vision-Language Modelscited by 19
  2. 2024Compositional 4D Dynamic Scenes Understanding with Physics Priors for Video Question Answeringcited by 18
  3. 2025Impossible Videoscited by 16
  4. 2025EgoNormia: Benchmarking Physical Social Norm Understandingcited by 13
  5. 2024Long Video Diffusion Generation with Segmented Cross-Attention and Content-Rich Video Data Curationcited by 12
  6. 2024Compositional Physical Reasoning of Objects and Events from Videoscited by 11
  7. 2024Physics Context Builders: A Modular Framework for Physical Reasoning in Vision-Language Modelscited by 8
  8. 2025Seeing is Not Reasoning: MVPBench for Graph-based Evaluation of Multi-path Visual Physical CoTcited by 6
  9. 2026What-If World: A Causal Benchmark for General World Models in Embodied Scenarioscited by 6
  10. 2025ENTER: Event Based Interpretable Reasoning for VideoQAcited by 5
  11. 2026PhysCodeBench: Benchmarking Physics-Aware Symbolic Simulation of 3D Scenes via Self-Corrective Multi-Agent Refinementcited by 5
  12. 2026Vision Language Models Cannot Reason About Physical Transformationcited by 4
13 more
  1. 2025Aligning Perception, Reasoning, Modeling and Interaction: A Survey on Physical AIcited by 3
  2. 2025InPhyRe Discovers: Large Multimodal Models Struggle in Inductive Physical Reasoningcited by 2
  3. 2026PhysEditWorld: A Large-Scale Dataset Toward Physics-Editable World Modelscited by 2
  4. 2025Agentic Physical AI toward a Domain-Specific Foundation Model for Energy Systems: A Case Study on Nuclear Reactor Controlcited by 1
  5. 2026Order from Chaos: Physical World Understanding from Glitchy Gameplay Videoscited by 1
  6. 2026PhysQuantAgent: An Inference Pipeline of Mass Estimation for Vision-Language Modelscited by 0
  7. 2026Beyond Static Vision: Scene Dynamic Field Unlocks Intuitive Physics Understanding in Multi-modal Large Language Modelscited by 0
  8. 2026FeynmanBench: Benchmarking Multimodal LLMs on Diagrammatic Physics Reasoningcited by 0
  9. 2026ChronoPhyBench: Do MLLMs Truly Understand the World or Merely Exploit Language Priors?cited by 0
  10. 2026SiPhy: Single-Image Physical Property Reasoningcited by 0
  11. 2026Teaching Vision-Language Models to Use the Scale They Are Given: Label-Free Equivariance Training for Metric Physical Reasoningcited by 0
  12. 2026Do World Models Make Better Robots? A Survey of Evaluation Benchmarks for Predictive Embodied Intelligencecited by 0
  13. 2026OmniFysics-Captioner Technical Report: Grounding Omni-Modal Understanding in the Physical World for Better Captioningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2022ComPhy: Compositional Physical Reasoning of Objects and Events from Videos
  2. 2026PhysMRV: Physical Memory Retrieval and Verification for Physics Plausibility Reasoning
  3. 2024LLMPhy: Parameter-Identifiable Physical Reasoning Combining Large Language Models and Physics Engines
  4. 2026PhysicsMind: Sim and Real Mechanics Benchmarking for Physical Reasoning and Prediction in Foundational VLMs and World Models
  5. 2023Intrinsic Physical Concepts Discovery with Object-Centric Predictive Models
  6. 2026Decoupled Physical Modeling and Execution for Physics Reasoning
  7. 2025PhyX: Does Your Model Have the "Wits" for Physical Reasoning?
  8. 2026PhyGround: Benchmarking Physical Reasoning in Generative World Models