WRSWorld Research StateContribute
WORLD RESEARCH STATE

Universal Physics Transformers: mesh-independent latent dynamics

Contribution grade

Neural operators for physical simulation · Human-designed research method or resource

Reference score80,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Encoded grid- or particle-based physical states into a fixed-size latent representation, propagated dynamics there with inverse encoding/decoding objectives and decoded at queried locations.

Scientific scope

Encoded grid- or particle-based physical states into a fixed-size latent representation, propagated dynamics there with inverse encoding/decoding objectives and decoded at queried locations. Universal denotes a shared architectural formulation across tested discretizations, not approximation/error certification for all PDEs, exact conservation or unrestricted spatial detail. Inherits neural operators/transformers; original JKU/NXAI affiliations are not retrospectively changed by a current company index.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Universal denotes a shared architectural formulation across tested discretizations, not approximation/error certification for all PDEs, exact conservation or unrestricted spatial detail. Inherits neural operators/transformers; original JKU/NXAI affiliations are not retrospectively changed by a current company index.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction