LOAIDBJan 14

Query Languages for Machine-Learning Models

arXiv:2601.09381v1h-index: 1STACS
Originality Synthesis-oriented
AI Analysis

This work addresses the problem of querying neural networks for researchers in logic and machine learning, but it appears incremental as it builds on existing logics from the 1990s.

The paper investigates first-order logic with summation (FO(SUM)) and its recursive extension IFP(SUM) as query languages for machine learning models like neural networks, represented as weighted graphs, by presenting examples and analyzing their expressiveness and computational complexity.

In this paper, I discuss two logics for weighted finite structures: first-order logic with summation (FO(SUM)) and its recursive extension IFP(SUM). These logics originate from foundational work by Grädel, Gurevich, and Meer in the 1990s. In recent joint work with Standke, Steegmans, and Van den Bussche, we have investigated these logics as query languages for machine learning models, specifically neural networks, which are naturally represented as weighted graphs. I present illustrative examples of queries to neural networks that can be expressed in these logics and discuss fundamental results on their expressiveness and computational complexity.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes