@Deliverator@kbin.social
@Deliverator@kbin.social avatar

Deliverator

@Deliverator@kbin.social

Tales from the Terrordrome

Este perfil es de un servidor federado y podría estar incompleto. Explorar más contenido en la instancia original.

BayesFlow: Amortized Bayesian Workflows With Neural Networks (arxiv.org) en

Modern Bayesian inference involves a mixture of computational techniques for estimating, validating, and drawing conclusions from probabilistic models as part of principled workflows for data analysis. Typical problems in Bayesian workflows are the approximation of intractable posterior distributions for diverse model types and...

Deliverator,
@Deliverator@kbin.social avatar

Frankly I think we need more people before we can start getting concerned about things like that. If we're trying to make the Fediverse a viable alternative it has to be appealing and easy enough to use that people want to use it. If we don't get that right this whole thing is doomed from the start

The Curse of Recursion: Training on Generated Data Makes Models Forget (arxiv.org) en

Stable Diffusion revolutionised image creation from descriptive text. GPT-2, GPT-3(.5) and GPT-4 demonstrated astonishing performance across a variety of language tasks. ChatGPT introduced such language models to the general public. It is now clear that large language models (LLMs) are here to stay, and will bring about drastic...

Deliverator,
@Deliverator@kbin.social avatar

Yeah it does not portend well for the future, especially combined with the current explosion of low quality, profit driven content. I fear if left unchecked we could approach some kind of Kessler Syndrome-style scenario where desire for rapid growth and profit will poison the well in the long term. "Garbage in, garbage out"

  • Todo
  • Suscrito
  • Moderado
  • Favoritos
  • •
  • random
  • noticiascr
  • CostaRica
  • Todos las revistas
    •