Work

Disambiguation of Notations: Ω, U, U₀, W-Weights

Year: 2026
Type: preprint
Abstract: At the intersection of MBM ontology (IDM — Infinite-Dimensional Multiverse) and machine learning (ML), a critical problem arises: identical letter designations begin to denote different concepts. The main conflict is associated with the letter W, which in fundamental ontology traditionally denotes the Universum (the union of all activated sets), while in ML it represents the neural network weight matrix. For applied AI tasks, such a mixture is unacceptable, as it leads to incorrect interpretation of formulas (e.g., W = ⋃_S A(S) is misread as a description of weights). This paper proposes a strict system of notations to eliminate this terminological ambiguity. The symbol U is firmly assigned to ontology to denote the Universum. The paper fixes the basic vocabulary of OI ontology (Ω, U, U₀, A(S), S, K, Id) and the vocabulary of machine learning (W, b, x, θ, s). For clarity, a summary table of correspondences is provided, delineating the meanings of symbols in the two disciplines. Four strict rules are introduced for all subsequent works of the OIE series: 1) the letter W always means weights only; 2) U₀ always means our Universe; 3) all ML terms must be explained at their first mention; 4) the concept of "structure" (K) cannot be denoted by weight matrices or parameters θ, as it describes topology. The adopted notation system creates a unified metalanguage for the OIE series, reduces the cognitive load on the reader, and ensures correct communication between ontologists and ML engineers.
Language: English
Open Access status: green
OpenAlex ID: W7237335192
Cites:
Cited by:
Related to: