Editorial approach
WAMPlatform is an independent research and discovery hub for World-Action Models, embodied AI and the technologies that enable them.
What we track
Original World-Action Model research, closely adjacent embodied-action systems, predictive world models, simulation, robot-learning methods and relevant enabling science. A paper can be pertinent without using the WAM label.
What a connection means
Links to a model’s original publication are distinguished from other source-backed technical relationships and editorial topic associations. A shared topic is not evidence of dependency, endorsement, citation or collaboration.
Source checking and limitations
Where noted, first-party sources support core model identity and broad method. Independent replication, benchmark claims, weights availability and commercial licensing are not implied. Newly discovered paper records may be indexed automatically from provider metadata; those entries are marked as unreviewed and should not be mistaken for scientific evaluations.
Rights and attribution
We link to source publications and identify authors. Automatically indexed records contain descriptive bibliographic metadata rather than reproduced full papers, figures or third-party abstracts. Rights in underlying research remain with their respective owners.
Corrections
Research metadata can change or be withdrawn. Please check the linked original publication for the authoritative version and use our contact page to request a correction once the contact channel is commissioned.
Research connections are an evolving editorial index, not a complete inventory of the field or a commercial endorsement.