Working notes from Ian's prompt on clarifying Causor language in Part 1.5, refining causal density definitions, and linking labeling/separability to prediction.
Definitions of computation, representation, information, and prediction within Epimechanics. Computation is producing a representation — discriminating, partitioning, compressing — and varies in depth from passive encoding to self-referential modeling. Connections to Shannon, Landauer, Constructor Theory, Church-Turing, Kolmogorov, and autopoiesis.
The seven foundational definitions of Epimechanics, when quotiented by their declared equivalences, converge to a four-node graph whose spectral radius is the golden ratio φ — a structural consequence of this definitional architecture.
Continuation from toy TP-bridge runs to a residual multi-layer architecture trace, preserving drift/update/transfer metrics and docs-ready visual outputs.
What the recent Tensor Programs bridge experiments are testing, how to read the charts, and what the initial results imply for Epimechanics.
