Černý reset bounds for positive-level one-cluster automata
Follow the latest progress by email.
For synchronizing one-cluster automata of positive level, prove the relative extending-word property and a reset word of length at most (m−1)(n−1)+mℓ≤(n−1)². The refined binary target is 3n−5 with attaining examples.
Why it matters and what counts as progress
Why it matters: It tests the Černý quadratic reset bound on a substantial structured automata class, with stronger class-specific bounds and sharpness.
What counts as progress: Prove the Černý reset bound for the full positive-level one-cluster class and establish sharpness of any stronger class-specific bound.
How the problem developed
- 22 Jul 2026 · Reported theorem [1]
Yinfeng Zhu, with OpenAI Codex (GPT-5.6 Sol) Claimed the Černý bound for positive-level synchronizing one-cluster automata, with reset length at most (m−1)(n−1)+mℓ ≤ (n−1)²; the author says the proof was verified by him.
AI contribution
The Černý Conjecture for One-Cluster Automata via Annular Spectral Descent
For synchronizing one-cluster automata of positive level, Zhu proves the relative extending-word property and reset bound (m−1)(n−1)+mℓ≤(n−1)^2. Binary examples attain his refined 3n−5 bound.
Sources
Website timeline
Publication dates and editorial changes.
Codex-assisted proof settles the positive-level one-cluster Černý bound