Fiber Fingerprints of Hidden Learning-State Dynamics

Abstract:A learning system can occupy execution states that are indistinguishable under every declared present-behavior readout yet respond differently to future training. We formalize this through fiber fingerprints: controlled future-learning response laws restricted to present-behavior equivalence classes. Prefix-compatible finite probes induce a predictive quotient functor, a Nerode-type minimal recursively sufficient representation, and a canonical set-level predictive fiber without assuming smoothness, reversibility, finite rank, or a manifold. Under an explicit finite-dimensional Hilbert realization, response decomposes into visible, visible-mode-reuse, and irreducible-new sectors; a history-reachability bridge retains only distinctions generated by natural training histories. Conditional mechanism results then identify a graph-Hodge chronology decomposition, a regular switching class with root-mean-square scale $\sqrt{p}\eta^{3/2}$ and finite-scale corrections, and an exact Adam moment section whose immediate adaptive field is constant while common future gradients can reveal hidden moment differences. Frozen Transformer--LoRA--AdamW studies with Qwen2.5-7B and Mistral-7B-v0.3 support a local action backbone, longer-horizon first-return non-closure, and fresh visible-relative completion with output-range reuse and a low-rank irreducible sector. Stronger claims remain bounded by preregistered negative or mixed results: re-anchored transport is unresolved above its measurement floor; the strict finite-grid Hodge--$3/2$ conjunction is unmet despite prospective contraction; Qwen accessibility is not established in the frozen raw moment chart; and Mistral revelation is future-context dependent rather than bank invariant. Within these support-, scale-, metric-, and context-resolved boundaries, present behavior is not a sufficient statistic for declared future learning.

30 pages, 8 figures, 7 tables. Ancillary files include figure-reproducibility code and frozen plot-level data

Machine Learning (cs.LG)

arXiv:2608.15976 [cs.LG]

(or arXiv:2608.15976v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2608.15976 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

README_SOURCE.md

figure_reproducibility/frozen_plot_data/README.md

figure_reproducibility/frozen_plot_data/completion_sectors.csv

figure_reproducibility/frozen_plot_data/plot_metadata.json

figure_reproducibility/frozen_plot_data/revelation_context.csv

figure_reproducibility/frozen_plot_data/revelation_scaling.csv

figure_reproducibility/frozen_plot_data/switching_exponents.csv

figure_reproducibility/make_data_figures.py

(3 additional files not shown) You must enabled JavaScript to view entire file list.

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.