This study investigates the continual pre-training of MoE transformers, revealing their robustness to distribution shifts and maintaining sample efficiency, outperforming dense models with lower costs.

https://arxiv.org/abs//2503.05029

YouTube: https://www.youtube.com/@ArxivPapers

TikTok: https://www.tiktok.com/@arxiv_papers

Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id1692476016

Spotify: https://podcasters.spotify.com/pod/show/arxiv-papers