Conjectured removal of the logarithmic factor in tensor robust PCA recovery

Let dd be the tensor order, let ndn_d denote the relevant tensor dimension, and consider the exact-recovery theorem for tensor robust PCA under the paper's stated assumptions. The assumption includes a factor lnmax{2d5,0}(nd)\ln^{\max\{2d-5,0\}}(n_d).

Tensor robust PCA recovery conjecture. The factor

lnmax{2d5,0}(nd)\ln^{\max\{2d-5,0\}}(n_d)

can be removed from the assumption without invalidating the exact-recovery conclusion of the theorem.

This conjecture seeks sharper conditions for exact recovery of tensors of arbitrary order, extending the matrix-case perspective. The paper establishes the recovery theorem with the logarithmic factor present, while its removal remains open.

Sources & referencesView supporting material

Primary source

Jiewen Guan, Bo Jiang and Zhening Li, “On decomposability and subdifferential of the tensor nuclear norm”, arXiv:2510.04647 (2026).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.