{"id":"arxiv/2608.05443","url":"https://arxiv.org/abs/2608.05443","tags":["math.LO","cs.IT","cs.LO"],"score":0,"title":"Sequential-Innovation Reducibility and the Innovation Spectrum","domain":"math.LO","source":"arxiv","authors":["Amir Leshem"],"reviews":0,"subject":"math.LO","summary":"Sequential prediction naturally induces an innovation sequence consisting of the prediction errors produced by a causal predictor. We use the collection of all such innovation sequences to define the \\emph{innovation spectrum} of an individual binary sequence and, from it, a new reducibility based on sequential information extraction. We show that this reducibility refines truth-table reducibility while exhibiting a fundamentally different geometry. The degree structure decomposes into two canonical regions: a truth-table spine, whose induced order is isomorphic to the truth-table degrees,…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-05","reactions":{}}