- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper embeds an -layer feedforward network in a dissipative-mechanics framework on a doubled phase space: activations and sensitivities are encoded as the mean and the difference (the stress) of a conjugate pair of global states and , and a single global Lagrangian is constructed by way of the Bateman–Galley formalism for non-conservative systems. At the unique stable equilibrium the exact BP gradient is given by the stress , and under unit-step Euler discretization BP is recovered exactly within steps; the algebraic mechanism is the strict lower block-triangularity of the global weight matrix , which makes satisfy . The experiments compare a 9-layer VGG on CIFAR-10 against PyTorch autodiff, with test accuracy on par with BP ($\sim$93%); among other readings, the relative gradient error and are reported; a block-grouping comparison is run on a reduced CNN (15/7/4 stages); and an inertia comparison on a linear recurrent model.
Critical strengths
- The paper writes the BP gradient as the stress at the unique stable equilibrium, and recovers standard BP in steps under unit-step Euler (pp. 1–2; Theorem 4.2 and Algorithm 2, pp. 6 and 16). (Lemma 3.1, p. 3) is the algebraic source of this finite-step conclusion; the authors report two practical consequences (p. 1, pp. 8–9): the number of convergence rounds for block grouping is exactly twice the number of stages (30/14/8), and the inertia result comes from a linear model with spectral radius close to 1. This writes the credit assignment of BP as an equilibrium expressed through the mean and difference of the conjugate states and gives an explicit -step recovery path, so that the physical formulation and the computational content of BP can be compared directly.
Critical weaknesses
- The paper does not connect the "physical formulation" to a usable capability: block grouping and inertia come only with simulation-level readings (pp. 8–9), the claimed analog/neuromorphic/hardware use case exists only as a design (Appendix G), and p. 9 states "hardware runtime has not been measured"; the Appendix G variant likewise reports no gradient or accuracy readings — the authors summarize it with "tested with identical accuracy" (p. 4) and "matches the performance of algorithm 1" (p. 18), but give no readings beyond the equations and Algorithm 3. The paper thus provides no checkable account of what this physicalized formulation of BP adds over existing physical descriptions of BP.
- The paper is hard to read and lacks both a concrete illustration of the concept and a positioning against related algorithms: the derivation proceeds through dense formulas, Figures 1–8 are all curves or heatmaps, and no figure maps and onto network units; beyond the comparison table in Section 6, the paper does not state how the formulation differs in computational content from existing algorithms such as PC/BP/EP.
Questions for authors
- The structure is already defined algebraically in §3.1 and Eq. (2); can the authors add a figure that maps the derived formulas onto the layer units, and state how the formulation differs from PC/BP/EP architectures (cf. Whittington et al., 2019)? And what does the formulation provide beyond what other algorithms already offer — where does its value lie (biological plausibility, training or inference cost, deployable hardware, or something else)?
- In §3.4, in Eq. (5) depends only on and ; taking its variation as written in Eq. (6) yields the equilibrium conditions of Eq. (10), while Eq. (7) carries , corresponding to the modified Euler–Lagrange equation, Eq. (26), in Appendix A. How does the main text get from Eq. (6) to Eq. (7)? Where does the Rayleigh term sit within the variational principle?
- Read literally, Eqs. (44)(45) and Algorithm 3 take the loss gradient at and rather than at the midpoint ; is the equilibrium so obtained different from BP? If "matches" refers to test accuracy, what is the per-layer gradient error of this variant?