Plasticity and Valence Reweighting over Irreversible Residues

This paper extends a previous formulation (CIM v2) that described how an irreversible residue within a self-maintaining boundary generates qualia, but did not explain how that residue is read out into valence and action weight. The extension defines transformation as a change in the metric of evaluation axes, using the irreversible residue as substrate, such that the metric does not fully return to its initial state. Valence and action weight are positioned as modulations read out over that altered metric. The claim is placed on the metric generating the distribution, not on the output distribution itself, with irreversibility grounded in the terminability of the self-maintaining boundary. The paper limits its scope to the readout layer, leaving the origin of qualia and the hard problem for future work.

Study at a glance

Abstract

CIM v2 (Ishikawa, 2026) formulated the irreversible residue that remains in a self-maintaining boundary as a condition for the generation of qualia, but it did not adequately give form to how that residue is read out into valence and action weight. This paper is an extension that fills this gap. First, it defines transformation (henshitsu) not as the fixed storage of experiential content but as a change—using the irreversible residue as substrate—in the very metric of the evaluation axes, such that the metric does not fully return to its initial state. Second, it positions valence and action weight as modulations read out over that altered metric. The locus of the claim is placed not on the output distribution itself but on the metric that generates the distribution, and its irreversibility is, within CIM, treated as grounded in the terminability of the self-maintaining boundary. This extension does not deny the theory of constructed emotion or engram research; rather, taking as given the history-dependence and plasticity they document, it asks under what conditions transformation is not reducible to a mere updating of the distribution. The scope of this paper is limited to the layer of readout; the origin of qualia and the hard problem itself are left as future work.