Said Slaoui

S-AI-RLM: A Recursive Logic Machine

S-AI-RLM introduces a Recursive Logic Machine within the Sparse Artificial Intelligence paradigm to address a central limitation of contemporary probabilistic reasoning systems: strong empirical performance does not by itself provide guarantees of termination, logical correctness...

S-AI-CCC: Controlled Cognitive Canonicalization

Background. Large language models routinely transform mathematical expressions, logical formulas, symbolic problems, and unstructured inputs into more regular, compressed, or conventionally normalized forms. We call this phenomenon implicit canonicalization: an emergent statistic...

S-AI-CCC: Controlled Cognitive Canonicalization

Background. Large language models routinely transform mathematical expressions, logical formulas, symbolic problems, and unstructured inputs into more regular, compressed, or conventionally normalized forms. We call this phenomenon implicit canonicalization: an emergent statistic...