TY  - RPRT
TI  - Efficient Training of Language Models with Compact and Consistent Next Token Distributions
AU  - Ashutosh Sathe
AU  - Sunita Sarawagi
PY  - 2024
UR  - https://arxiv.org/abs/2407.02819
ID  - 2407.02819
ER  -