Metraipolis › archive › dc_7d1e5f8a2e62

Relevance Propagation as Entropic Optimal Transport on Filippov Neural Networks: Sinkhorn Propagation and Region Relevance in Image Classifiers — arXiv:2611.05594

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.05594
License (as archived)
CC0-1.0
Archived
2026-11-05T18:21:56.700Z
SHA-256 of the archived text
6eba3231c0770e712c80b69fb9dd54f9ae6ee241d350c33ae929a7e8c49f29a9
Stable URL of this page
https://metraipolis.com/documento/dc_7d1e5f8a2e62

How to cite

Relevance Propagation as Entropic Optimal Transport on Filippov Neural Networks: Sinkhorn Propagation and Region Relevance in Image Classifiers — arXiv:2611.05594 — arXiv (en), https://arxiv.org/abs/2611.05594 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7d1e5f8a2e62 (Metraipolis archive dc_7d1e5f8a2e62, sha256 6eba3231c0770e712c80b69fb9dd54f9ae6ee241d350c33ae929a7e8c49f29a9)
BibTeX
@misc{dc7d1e5f8a2e62,
  title = {Relevance Propagation as Entropic Optimal Transport on Filippov Neural Networks: Sinkhorn Propagation and Region Relevance in Image Classifiers — arXiv:2611.05594},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.05594},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7d1e5f8a2e62; sha256 6eba3231c0770e712c80b69fb9dd54f9ae6ee241d350c33ae929a7e8c49f29a9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.05594. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Relevance propagation, which redistributes a network's prediction backward through the layers, is usually treated as a heuristic. We show that it is the solution of a single variational problem: Layer-wise Relevance Propagation (LRP) is the divergence-free flow on the active computational graph of zero relative entropy to the backward transition law of the forward activation flow; for a non-negative reference, this is a relative entropy between Markov measures on backward paths. The rules LRP$_0$, LRP$_{z^+}$ and LRP$_{\alpha,\beta}$ follow from this principle by the choice of a reference, and are instances of one rule, LRPN$_{\alpha,\beta}$, whose weights may vary from node to node, for instance with the node's activity, in the spirit of Hebb's principle. The principle rests on a measure-theoretic formalization of feed-forward networks as directed acyclic graphs of Filippov operators, which covers piecewise-smooth and discontinuous architectures (e.g. hard-routed Mixture-of-Experts) with a Clarke-Filippov generalized-gradient calculus; it also characterizes backpropagation as the adjoint equation of a Lagrangian relaxation. On networks of ReLU and max-pooling nodes, LRP$_0$ is Gradient$\times$Input, with an explicit jump across a kink where LRP$_{z^+}$ stays continuous. Taking the entropy edge by edge yields a new rule, Sinkhorn Relevance Propagation. Aggregated over an input region, the flow of a non-negative reference tells, with exact accounting, which hidden units carry its evidence, the conditioned flow being a Schrödinger bridge. On VGG-16, the conservation identities hold to round-off, and the rules are told apart by the sign of their maps, the share of the output left to the bias nodes, faithfulness, localization and class dependence; on ConvNeXt-Tiny, LRPN$_{\alpha,\beta}$ conserves the output and its Hebbian weights give the highest ABPC of the rules tested.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Computer Vision and Pattern Recognition (cs.CV); Optimization and Control (math.OC).
Pagina del lavoro: https://arxiv.org/abs/2611.05594 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-05T18:15:49.001Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).