duccio

duccio

@duccio.bsky.social

37 followers114 following6 posts

duccio reposted

๐Ÿšจ First ever post here~ New paper out! Instruct models are not always the best. Scaling down ๐Ÿ“‰ instruction tuning strength via partial adaptation leads to material gains ๐Ÿš€ on few-shot in-context learning NLP tasks across model families and sizes. arxiv.org/abs/2504.11626

You are up to date.