Human-Centered Learning Mechanics: A Dynamical Framework for Entropy-Regulated Representation Learning
For researchers in deep learning theory, this work offers a formal framework to understand when entropy regularization is effective, but it is incremental as it builds on existing dynamical and information-theoretic perspectives without introducing a new paradigm or achieving SOTA results.
The paper proposes Human-Centered Learning Mechanics (HCLM), a dynamical framework for entropy-regularized representation learning, and shows that geometric entropy surrogates (especially log-determinant covariance) induce stronger and more stable information forces than softmax-normalized entropy, with controlled experiments supporting the hypothesis.
Deep learning is increasingly viewed as a dynamical process in parameter space, yet many existing theories still treat training as a closed optimization system. This view is limited for real-world AI, where models operate under uncertainty, resource constraints, distribution shift, downstream decision risks, and human feedback. We propose Human-Centered Learning Mechanics (HCLM), a dynamical and information-theoretic framework for open and controlled learning systems. The central idea is that entropy regularization is useful only when the chosen entropy surrogate generates a non-degenerate information force along the optimization trajectory. Otherwise, entropy terms may produce weak, unstable, or misaligned gradients, causing the dynamics to collapse toward ordinary loss minimization. We introduce the notion of effective entropy and study tractable geometric entropy surrogates, including variance-based and log-determinant covariance proxies. The paper makes three contributions. First, it formalizes entropy regularization through effective information force and characterizes degenerate entropy regimes. Second, it derives convergence, entropy-flow, Wasserstein-gradient-flow, and noisy-representation generalization results under explicit assumptions. Third, it offers a conditional dynamical interpretation of scaling-law-like behavior as a balance between information injection, entropy dissipation, and residual risk, without claiming an unconditional derivation of empirical neural scaling laws. Controlled representation-learning experiments support the hypothesis that geometric entropy surrogates, especially log-determinant covariance entropy, induce stronger and more stable information forces than softmax-normalized entropy.