It is shown that meta-learning can substantially reduce or entirely eliminate performance differences across architectures and data representations, suggesting that these factors may be less important as sources of inductive bias than is typically assumed.
A curious phenomenon called mode connectivity, the ability to connect neural networks in the loss surface, defies explanation entirely is elucidates, explains and exploits this special structure in the loss landscape.
This paper introduces the idea of {\em magnitude potential} -- a quantity based on the theory of metric magnitude -- that reflects how well an arbitrary point is represented by a given set, and finds that this basic quantity can be applied to examine various features in neural generalization.
Sahel Torkamani, Henry Gouk, Rik Sarkar· 0 citations
Artificial neural networks (ANNs) are computational models inspired by the way biological nervous systems process information. Over the past eight decades they have grown from simple threshold units into deep, billion-parameter systems that underpin modern computer vision, language processing, healthcare analytics, fin...
Megha D. Angnani S. B. Bele, Astha V. Thakare Ishwari P. Telkhade· International Journal of Adv...· 0 citations
Neuroscientific learning theories propose that the brain acquires knowledge by constructing internal world models. Supervised learning, the dominant approach in deep neural networks (DNN), relies on external category labels, making it difficult to reconcile with biological learning. There is an increasing trend towards...
Dirk Gütlin, Denise Kittelmann, Ryszard Auksztulewicz· bioRxiv· 0 citations
