Obrázek epizody DSB Podcast #25 [CZ] - Softmax fix

DSB Podcast #25 [CZ] - Softmax fix

Obrázek epizody DSB Podcast #25 [CZ] - Softmax fix

Poslechněte si podcast

Dnes

37 min

In this episode, hosts delve into a technical discussion of a recent paper by DeepMind titled "Fix for Softmax and Cross Entropy," which tackles the numerical stability issues in neural networks. The hosts explore the implications of the paper's findings, particularly its potential to speed up learning processes by correcting numerical errors inherent in floating point operations. Additionally, the episode covers the upcoming IPO of Anthropic, highlighting the financial complexities and potential risks associated with their ambitious growth and dependency on major tech investors like Nvidia, Microsoft, and Google.


Link to the blogpost

Popis podcastu

Data Science Bulletin is a podcast about news in space of data science, artificial intelligence and machine learning. We bring you views of people "on the ground", implementing these methods into industry setting. Podcast in Czech, accompanying posts and newsletter in English.