Hierarchical Per-Module Gradient Calibration Resolves Interference in Multi-Hop Rule-Routing AttributionGlobal query standardization and module weighting over summed projected ($d_{\mathrm{proj}}=16$) negative log-odds (`NLO`) gradients raised CATT `Rule Routings`
Query-Calibrated Token Gradients and Learned Module Weighting for Rule-Routing AttributionLow-rank projected gradient attribution on CATT performs strongly on direct constant-binding memorization (`Function Bindings with distractors`, `FB-distr`) yet