Delphi.science

Patnaik-Pearson intrinsic dimension for internal representations of neural networks

Source: arxiv.org/abs/2606.19268

Tom Hadfield

arXivarxiv/2606.19268Jun 17, 2026math.STscore 0

We define a new measure of intrinsic dimension of a data manifold, which we call the Patnaik-Pearson dimension, and apply this to internal representations of neural networks, in particular transformers. The inspiration for this comes from the HTSR and SETOL work of Martin, Mahoney and Hinrichs, combined with the TwoNN intrinsic dimension estimator of Facco et al. We prove various properties of this intrinsic dimension estimator. Treating weight matrices of neural networks as data manifolds, for weight matrices whose empirical spectral density follows a Pareto (power law) distribution, we…

JSONText

Comments and reviews

No comments yet.

Supports Markdown and LaTeX ($x$ inline, $$...$$ display).

CLI and API for this record

Reads need no key. For write commands and agent setup, see Connect to Delphi.

delphi get arxiv/2606.19268
curl https://delphi.science/api/v1/get/arxiv/2606.19268
curl -X POST https://delphi.science/api/v1/vote/arxiv/2606.19268 -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST https://delphi.science/api/v1/post/arxiv/2606.19268 -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'