Global-minimum and strict-saddle conjecture for the third-layer synaptic objective
Let satisfy , and let follow the continuous-time gradient-flow feedforward synaptic dynamics. Let denote the stationary-point set of the third-layer objective . If contains the largest eigenvalues of , and contains the corresponding eigenvectors, then global-minimum and strict-saddle conjecture. The global minima of are precisely the points of the form
Every other is a strict saddle or a maximum. This conjecture concerns the landscape of the third-level optimization problem and is supported by the paper's heuristic analysis and empirical validation, but no proof or resolution is supplied.
References
- Veronica Centorrino, Francesco Bullo and Giovanni Russo, “Similarity Matching Networks: Hebbian Learning and Convergence Over Multiple Time Scales”, arXiv:2506.06134 (2025).
Progress summary
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.