Metraipolis › archive › dc_74bfd4970b08

Certified Approximation for Interpretable Representer Landmarks — arXiv:2609.38901

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.38901
License (as archived)
CC0-1.0
Archived
2026-10-07T13:45:37.097Z
SHA-256 of the archived text
15b3dc6884a3125054c6565eb1e6f59c1c37aa28121ccacd8150685ba7c1aac1
Stable URL of this page
https://metraipolis.com/documento/dc_74bfd4970b08

How to cite

Certified Approximation for Interpretable Representer Landmarks — arXiv:2609.38901 — arXiv (en), https://arxiv.org/abs/2609.38901 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_74bfd4970b08 (Metraipolis archive dc_74bfd4970b08, sha256 15b3dc6884a3125054c6565eb1e6f59c1c37aa28121ccacd8150685ba7c1aac1)
BibTeX
@misc{dc74bfd4970b08,
  title = {Certified Approximation for Interpretable Representer Landmarks — arXiv:2609.38901},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.38901},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_74bfd4970b08; sha256 15b3dc6884a3125054c6565eb1e6f59c1c37aa28121ccacd8150685ba7c1aac1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38901. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Representer explanations rank the training landmarks that most influence a self-supervised representation. At scale, this ranking rests on up to four stacked approximations of the empirical neural tangent kernel (eNTK). These are random output heads, a parameter sketch, landmark sampling and a coefficient fit. Existing analyses bound each approximation separately, but none certifies the top-$K$ set against their combined error. We introduce CAIRN (Certified Approximation for Interpretable Representer laNdmarks), a framework that carries this error through to the ranking. We derive the exact variance of the sketched multi-head eNTK, which matches measurement within $4\%$ where Johnson-Lindenstrauss bounds err by up to $2.5\times$. This yields a high-probability top-$K$ certificate for a fixed coefficient fit, alongside exact residual-trace certificates for discarded spectral mass. An exact product-variance identity separates kernel error from fit variability and identifies when a larger kernel budget can still sharpen a ranking. Stochastic Lanczos Quadrature (SLQ) estimates the effective dimension within $0.72\%$ and guides the landmark budget without dense eigendecomposition. We show that residual mass does not control class coverage, and residual-greedy selection cuts the worst coverage excess of $k$-means++ from $8.5\times$ to $1.55\times$ ($4\times$ on the sketched eNTK). Cross-view initializers outperform principal-component initialization in five (AUI) to all six (CSI) settings. Against the KREPES Gauss-Newton solver, CAIRN converges $2.5$ to $11.3\times$ faster, trails by at most $0.31$ points and gains up to $3.14$ points on MNIST. Together, these results make the reliability of representer explanations measurable and show where approximation budgets are best spent.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.38901 — 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 2026-10-07T13:18:10.279Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).