Lacuna Inc. presents the Invariant-Variant Disentangled State-Space Model (IVD-SSM) as its submission to SemEval-2026 Task 4 on Narrative Story Similarity and Narrative Representation Learning. The model leverages a hybrid State-Space Model backbone, Jamba-1.5-Mini, to model extended causal chains without the quadratic bottlenecks of standard Transformers. It introduces a Structurally Gated Alignment (SGA) head that maps a coarse structural skeleton via a Macro-path to gate a full-resolution Micro-path, suppressing semantic noise and superficial keyword overlaps. Evaluated on pairwise comparative judgments (Track A) and dense representation learning (Track B), the approach demonstrates that disentangling structural invariants from lexical variants provides a robust framework for deep narrative understanding.