Multi-Scale Temporal Homeostasis Enables Efficient and Robust Neural Networks
MD. Azizul Hakim · 2026
Read Multi-Scale Temporal Homeostasis Enables Efficient and Robust Neural Networks by MD. Azizul Hakim (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-2602-07009.txt.
Description
MD Azizul Hakim builds a framework that borrows a trick from living brains: regulation on many time scales at once. Artificial nets score well on tests yet break under small shocks. Living nervous systems, by contrast, stay reliable for decades, thanks to homeostatic control spread across time scales. His framework, MSTH, weaves four scales into artificial nets: ultra-fast, fast, medium, and slow.
The point is not surface mimicry. The scales coordinate as one hierarchy, using optimization tricks refined by evolution. Tests on molecular, graph, and image sorting tasks show steady gains in score. Catastrophic failures vanish, and recovery from shocks speeds up. The framework beats both single-scale bio-inspired rivals and top standard methods, which suggests cross-scale coordination is a core stabilizer for artificial nets.
Catalog record
| CREATOR | MD. Azizul Hakim |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-07009