Publication
Complexity Control by Gradient Descent in Deep Networks. Nature Communications 11, (2020).
s41467-020-14663-9.pdf (431.68 KB)
Compositional sparsity of learnable functions. Bulletin of the American Mathematical Society 61, 438-456 (2024).
Theory I: Deep networks and the curse of dimensionality. Bulletin of the Polish Academy of Sciences: Technical Sciences 66, (2018).
02_761-774_00966_Bpast.No_.66-6_28.12.18_K1.pdf (1.18 MB)
Is Research in Intelligence an Existential Risk?. (2014).
Is Research in Intelligence an Existential Risk.pdf (571.42 KB)
On efficiently computable functions, deep networks and sparse compositionality. (2025).
Deep_sparse_networks_approximate_efficiently_computable_functions.pdf (223.15 KB)
Theory II: Landscape of the Empirical Risk in Deep Learning. (2017).
CBMM Memo 066_1703.09833v2.pdf (5.56 MB)
How Deep Sparse Networks Avoid the Curse of Dimensionality: Efficiently Computable Functions are Compositionally Sparse. (2022).
v1.0 (984.15 KB)
v5.7 adding in context learning etc (1.16 MB)
Explicit regularization and implicit bias in deep network classifiers trained with the square loss. arXiv (2020). at <https://arxiv.org/abs/2101.00072>
Why and when can deep-but not shallow-networks avoid the curse of dimensionality: A review. International Journal of Automation and Computing 1-17 (2017). doi:10.1007/s11633-017-1054-2
art%3A10.1007%2Fs11633-017-1054-2.pdf (1.68 MB)
From Marr’s Vision to the Problem of Human Intelligence. (2021).
CBMM-Memo-118.pdf (362.19 KB)
From Associative Memories to Powerful Machines. (2021).
v1.0 (1.01 MB)
v1.3Section added August 6 on self attention (3.9 MB)
Turing++ Questions: A Test for the Science of (Human) Intelligence. AI Magazine 37 , 73-77 (2016).
Turing_Plus_Questions.pdf (424.91 KB)
Theory I: Why and When Can Deep Networks Avoid the Curse of Dimensionality?. (2016).
CBMM-Memo-058v1.pdf (2.42 MB)
CBMM-Memo-058v5.pdf (2.45 MB)
CBMM-Memo-058-v6.pdf (2.74 MB)
Proposition 4 has been deleted (2.75 MB)
What if.. (2015).
What if.pdf (2.09 MB)
An Overview of Some Issues in the Theory of Deep Networks. IEEJ Transactions on Electrical and Electronic Engineering 15, 1560 - 1571 (2020).
I-theory on depth vs width: hierarchical function composition. (2015).
cbmm_memo_041.pdf (1.18 MB)
The History of Neuroscience in Autobiography Volume 8 8, (Society for Neuroscience, 2014).
Volume Introduction and Preface (232.8 KB)
TomasoPoggio.pdf (1.43 MB)
Theoretical Issues in Deep Networks. (2019).
CBMM Memo 100 v1 (1.71 MB)
CBMM Memo 100 v3 (8/25/2019) (1.31 MB)
CBMM Memo 100 v4 (11/19/2019) (1008.23 KB)
Theoretical issues in deep networks. Proceedings of the National Academy of Sciences 201907369 (2020). doi:10.1073/pnas.1907369117
PNASlast.pdf (915.3 KB)
Deep Learning: mathematics and neuroscience. (2016).
Deep Learning- mathematics and neuroscience.pdf (1.25 MB)
Compositional Sparsity of Learnable Functions. (2024).
This is an update of the AMS paper (230.72 KB)
Visual Cortex and Deep Networks: Learning Invariant Representations. 136 (The MIT Press, 2016). at <https://mitpress.mit.edu/books/visual-cortex-and-deep-networks>
Loss landscape: SGD has a better view. (2020).
CBMM-Memo-107.pdf (1.03 MB)
Typos and small edits, ver11 (955.08 KB)
Small edits, corrected Hessian for spurious case (337.19 KB)
Deep Leaning: Mathematics and Neuroscience. A Sponsored Supplement to Science Brain-Inspired intelligent robotics: The intersection of robotics and neuroscience, 9-12 (2016).
Cervelli menti algoritmi. 272 (Sperling & Kupfer, 2023). at <https://www.sperling.it/libri/cervelli-menti-algoritmi-marco-magrini>
]