CNT Cross-Domain Decoder v1 — Sign-Aware Reconstruction of Domains from Multi-Domain Drift Fields

OverviewThis dataset provides the first Cross-Domain Decoder release for Cognitive Nexus Theory (CNT): CNT Cross-Domain Decoder v1 (CDDT v1). It quantifies how well one domain’s behaviour can be decoded from the others using CNT drift/hazard fields and related features, across finance, crypto, volatility, and other domains used in the CNT backbone. In plain language: given the CNT field, how well can we reconstruct “what SPY is doing” from {BTC, VIX, …}, or “what VIX is doing” from the others? CDDT v1 is intended as a companion to: CNT Cross-Domain Hazard Pure v1 (baseline hazard planes), CNT Global Drift Field v1 (the D_global index), and CNT Cross-Domain Echo Window v1 (lead–lag “echo windows”). Where those datasets describe the geometry and timing of drift, CDDT v1 tests whether that shared geometry is informational: whether one system can be reliably decoded from the rest. Files in this release (CNT_CDDT_v1_release.zip) CDDT_v1_README.txt Human-readable documentation for this release: how the decoder was set up, which domains and targets are included, and how to interpret each matrix and summary file. cnt_cddt_auc_matrix.csv Matrix of decoding performance (e.g. AUC values) for all source → target domain pairs. Each cell summarises how well the decoder predicts the target domain from the remaining domains under the chosen setup (see README for exact details). cnt_cddt_pval_matrix.csv Matrix of p-values comparing observed decoding performance against a null or surrogate baseline, indicating which source→target decoders carry statistically significant information beyond chance (see README for the exact null model). cnt_cddt_signaware_auc_matrix.csv “Sign-aware” decoding performance matrix: a variant that takes into account not just the magnitude but the direction (sign) of the target behaviour, useful for distinguishing “gets the side of the move right” from “just segments the regime.” cnt_cddt_summary_pairs.csv Flattened summary table of decoder results for each source→target pair, including AUC and possibly other metrics/flags. cnt_cddt_signaware_summary_pairs.csv Sign-aware version of the pairwise summary, focusing on sign-sensitive decoding performance. Together, these files provide a compact view of who can decode whom, and with what strength, in the CNT field. Role in Cognitive Nexus Theory (CNT)CDDT v1 plays three roles in the CNT ecosystem: Information test: It tests whether drift/hazard features across domains truly share structure, by asking if one domain can be reconstructed from the rest. Complement to echoes: It complements the Cross-Domain Echo Window dataset: echoes say “X tends to move before Y at lag H”, while the decoder asks “can we recover Y from the entire field?” Backbone check: It serves as a consistency check on the hazard and drift backbones, probing whether they encode meaningful cross-domain information or just cosmetic correlations. Intended useThis dataset is suitable for: Analysing which domains are easiest or hardest to decode from others. Comparing decoding patterns against echo windows and Forbidden Drift Zones. Testing alternative decoder architectures or null models using the same matrices as benchmarks. Exploring whether certain domains act as “information hubs” (good sources) or “opaque nodes” (hard targets). License and reuseThis dataset is released to enable open replication, criticism, and extension. Please cite this Zenodo record and related CNT backbone datasets when using it.