Brain-inspired sparse training enables Transformers and LLMs to perform as fully connected

Add code
Jan 31, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: