Neural Fields as World Models
Joshua Nunley ([email protected]) · 2026
Read Neural Fields as World Models by Joshua Nunley ([email protected]) (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-18690.txt.
Description
People can rehearse a catch in their heads before they move, which suggests the brain keeps a map-like model of the world rather than a bare list of states. Joshua Nunley, writing alone in 2026, proposes world models that preserve that spatial layout, so predicting physics becomes motion across a map instead of a jump between abstract states.
The implementation uses motor-gated neural fields, where local links spread activity and motor commands scale selected channels. In three tests the same design predicted ballistic motion without teleporting, improved a catching policy offline by passing error through a frozen model, and grew body-selective channels without body labels. The work is early evidence for one shared basis behind prediction, offline learning, and body sense.
Catalog record
| CREATOR | Joshua Nunley ([email protected]) |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-18690