Tensors of Ostensive Definitions: A Metric Framework for XAI and Invention Generation

Read full text Summarise with VIRAT
Authors: Mykola Dubovykov
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract. A neural network can make a diagnosis — but cannot explain why; this makes its decision unacceptable wherever a legally sound and reproducible result is required. We propose a formalism in which every feature is a direct physical measurement (an ostensive predicate) and the complete state space T is generated combinatorially from n such parameters via the formula ∑C(n,m) = 2ⁿ; a weighted Hamming metric δ and a realizability measure ρ* are introduced to assess the potential of unexplored hypothesis nodes in T. The formalism is demonstrated on a system for identifying dangerous diseases (n = 3, 15 nosological entities including COVID-19) and the trademark search system PromSearch — showing applicability from medical diagnostics to robotics, in all cases guaranteeing full explainability and verifiability of decisions.

Fields of Study & Subjects

Explainable Artificial Intelligence (XAI) Artificial Intelligence in Healthcare Diverse Interdisciplinary Research Studies

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
22 March 2026
Language
EN