We are proud to announce that Analyzing Differential Fuzzy Logic Operators, Emile van Krieken, Erman Acar and Frank van Harmelen is published in Artificial Intelligence Journal.

The paper presents an extensive analysis on the behaviour of a large collection of fuzzy logical operators in a differentiable learning setting. It points out a phenomenon: a strong imbalance between gradients driven by the antecedent and the consequent of the implication. In order to tackle this, it introduces a new family of fuzzy implications (called sigmoidal implications). It turns out that a non-standard combinations of logical operators which perform best in learning, yet surprisingly it seems they no longer satisfy the usual logical laws!