S. Gunasekar, J. Lee, D. Soudry, и N. Srebro. (2018)cite arxiv:1802.08246Comment: (1) A bug in the proof of implicit bias for matrix factorization was fixed. v2 gives a characterization of the asymptotic bias of the factor matrices, while v1 made a stronger claim on the limit direction of the unfactored matrix. (2) v2 also includes new results on implicit bias of mirror descent with realizable affine constraints.
K. Kawaguchi, L. Kaelbling, и Y. Bengio. (2017)cite arxiv:1710.05468Comment: To appear in Mathematics of Deep Learning, Cambridge University Press. All previous results remain unchanged.