Component attribution and latent feature count in HRM
Determine the causal contributions of HRM components (hierarchical recursion, deep supervision, ACT, and the choice of latent features) to performance, and justify the use of exactly two latent features versus alternative configurations.
References
Given the lack of ablation table in their paper, the over-reliance on biological arguments and fixed-point theorems (that are not perfectly applicable), it is hard to determine what parts of HRM is helping what and why. Furthermore, it is not clear why they use two latent features rather than other combinations of features.
— Less is More: Recursive Reasoning with Tiny Networks
(Jolicoeur-Martineau, 6 Oct 2025) in Section “Hierarchical interpretation based on complex biological arguments”