VibeMathedMath problems solved by AI

Worst-Case Complexity of Shellsort with Tokuda's Gap Sequence

Shellsort's worst-case running time is unknown for the gap sequences actually used in practice. Encoding a permutation as the polynomial σ(1)z++σ(n)zn\sigma(1)z + \cdots + \sigma(n)z^n gives a framework for lower bounds, and yields Ω(N1.26)\Omega(N^{1.26}) for Tokuda's 1992 sequence, extending to any strictly decreasing sequence staying within a fixed distance of a rational geometric one.

Result
Proved(see note)
Status
Partial result
AI contribution
AI-assisted
Method
Argument
Field
Analysis of algorithms
Posed by
Donald Shell; Naoyuki Tokuda
Year posed
1992
Years open
34y
Solved
2026-07-10
Model
ChatGPT
Vendor
OpenAI
Collaborators
Zhenghan Zang
Verification
Unreviewed
Publication
Preprint
Significance
15 / 100
Disclosed cost
Wikipedia
No dedicated article

What was actually shown

a lower bound for Tokuda's sequence; the general Shellsort complexity question stays open

What the AI did

The acknowledgement credits the model with providing the initial framework of the proof of Lemma 2, which the author then verified, refined and wrote up.

Verification

Single-author arXiv preprint; not yet peer-reviewed.

Source

arXiv:2607.08997 - Improved lower bounds of the time complexity of shellsort

Discussion