Category-Native Solomonoff Approximation: From Algorithmic Geometry to Kernels, Operators, and Induction †
Boumediene Hamzi, Marcus Hutter · 2026
Read Category-Native Solomonoff Approximation: From Algorithmic Geometry to Kernels, Operators, and Induction † by Boumediene Hamzi, Marcus Hutter (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/pmc-13605386.txt.
Description
Boumediene Hamzi and Marcus Hutter set out a 2026 theory paper on how to copy a classic ideal learner in practice. The ideal learner weighs all programs that could make the data, by their length. No real computer can run it. The authors argue that any workable copy must first pick the math world it will live in.
Each world brings its own way to compare, to code cost, and to predict. The paper builds out one branch in full, based on spaces with distance and mass. It proves steadiness and test-set bounds there. Other branches, such as graphs and trees, are laid out as a testable program. Gains and losses at each step are stated in plain view.
Catalog record
| CREATOR | Boumediene Hamzi, Marcus Hutter |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 229 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |