Eigenvalue Crossings and Discontinuous BBP Transitions in Spiked Euclidean Random Matrices
Source: arxiv.org/abs/2611.15927
We study the recovery of a rank-one signal embedded in a symmetric Euclidean random matrix, whose dense and geometrically correlated entries generally lack rotational invariance. We show that signal recovery through the leading eigenvector is governed by a competition between the principal directions of the noise and the signal. In particular, the signal may already be encoded in a subleading eigendirection, while remaining undetected by the leading principal component. As the signal strength increases, a crossing between competing eigenvalues can change their ordering, causing the leading…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2611.15927
curl -X POST /api/v1/records/arxiv/2611.15927/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2611.15927/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'