ML Reproducibility Challenge 2021 - Does Self-Supervision Always Improve Few-Shot Learning?

Does Self-Supervision Always Improve Few-Shot Learning? In this blog, Arjun provides a sneak-peek into his team's journey to fully reproduce a state-of-the-art paper for self-supervised learning based on few-shot learners, as part of the ML Reproducibility Challenge by Papers with Code.

challenge few-shot learning learning ml open source data science reproducibility