arxiv.org/pdf/1805.08522

N/A

# link_12qr17hk9mx.pdf ## Metadata - PDFFormatVersion=1.5 - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - CreationDate=D:20190423011718Z - Creator=LaTeX with hyperref package - ModDate=D:20190423011718Z - Custom.PTEX.Fullbanner=This is pdfTeX, Version 3.14159265-2.6-1.40.17 (TeX Live 2016) kpathsea version 6.2.2 - Producer=pdfTeX-1.40.17 - Trapped=False ## Contents ### Page 1 Published as a conference paper at ICLR 2019DEEP LEARNING GENERALIZES BECAUSE THE PARAMETER-FUNCTION MAP IS BIASED TOWARDS SIMPLE FUNCTIONSG

saved by

• Yudhister Joel Kumar
• Lev Chizhov

related reading

• Bayesian Neural Networkscs.toronto.edu
• nn-notes.pdfboris-hanin.github.io
• [1805.08522] Deep learning generalizes because the parameter-function map is biased towards simple functionsarxiv.org
• [2503.02113] Deep Learning is Not So Mysterious or Differentarxiv.org
• Understanding deep learning requires rethinking generalizationarxiv.org
• The generalization phase diagram — LessWronglesswrong.com
• Maybe I was too harsh on deep learning theory (three days ago) — LessWronglesswrong.com
• Some Math behind Neural Tangent Kernel | Lil'Loglilianweng.github.io
• Deep Learning is Not So Mysterious or Differentarxiv.org
• A Theory of Deep Learning | Elements of a Vector Spaceelonlit.com
• [1912.02178] Fantastic Generalization Measures and Where to Find Themarxiv.org
• [1802.05296] Stronger generalization bounds for deep nets via a compression approacharxiv.org