physical-reasoning

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

Recent Papers including Neural Symbolic Reasoning, Logical Reasoning, Visual Reasoning, planning and any other topics connecting deep learning and reasoning

Repo for "Large Language Model Reasoning Failures"

Phy-Q: A Testbed for Physical Reasoning

Learning Physics-grounded 4d Dynamics With Neural Gaussian Force Fields

📄 A curated list of visual reasoning papers.

[TMLR'24] A CommonSense Reasoning Dataset pertaining to Physical Commonsense affordance of objects.

PAWBench evaluates whether video generators recover the distribution of possible physical futures through repeated rollouts.

A Python toolkit for physical reasoning in LLMs and VLMs. This toolkit streamlines access to various physical reasoning datasets/benchmarks and provides a unified interface to different LLM/VLM providers, enabling researchers and developers to analyze, evaluate, and build applications in the physics reasoning domain.

A curated list of code world models: executable dynamics, programmatic simulation, and coding agents that build and reason about worlds.

Learning To Plan Tasks with Generalizable and Rapid Physical Reasoning for Embodied Manipulation

AI-powered VLM physical-world reasoning app using Knowledge Graph retrieval, physics rules, Firebase AI, web context, and ablation evaluation.

Project website for PAWBench: How Far Are We from Probabilistically Aligned World Modeling?

Framework for auditing the physical state-transition commitments of vision-language models. Instead of scoring only the answer, it elicits a typed trace, verifies it with a hybrid checker, and surfaces hidden inconsistency. Includes WMW-TRACEBANK: 200 validated traces across 17 physics families + 3,200 preference pairs.

A novice-first spaced-repetition deck for physical reasoning and measurement.

Neural Gaussian Force Field to generate 4D physics-grounded dynamics.

面向三维人体—场景交互的可验证物理推理：融合 SMPL-X、ScanNet、关系感知 Grounder 与 Qwen 证据生成。