2 Matching Annotations
  1. Jun 2023
  2. cdn.openai.com cdn.openai.com
    Language Models are Unsupervised Multitask Learners
    1
    1. mark.crowley 28 Jun 2023
      Recent work in computer vision has shown that common im-age datasets contain a non-trivial amount of near-duplicateimages. For instance CIFAR-10 has 3.3% overlap betweentrain and test images (Barz & Denzler, 2019). This results inan over-reporting of the generalization performance of ma-chine learning systems.

      CIFAR-10 performance results are overestimates since some of the training data is essentially in the test set.

      image-processing convolutional-neural-networks deep-learning machine-learning datasets
    Visit annotations in context

    Tags

    Annotators

    URL

    cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf
  3. Aug 2017
  4. arxiv.org arxiv.org
    1708.04347.pdf
    1
    1. hodapp 31 Aug 2017

      This is a very easy paper to follow, but it looks like their methodology is a simple way to improve performance on limited data. I'm curious how well this is reproduced elsewhere.

      convolutional neural networks neural networks deep learning
    Visit annotations in context

    Tags

    Annotators

    URL

    arxiv.org/pdf/1708.04347.pdf