Research papers improving performance of LLMs [3/3]

Research papers published from January 16th to February 15th, 2025 proposing context length and architectural changes in LLMs

In partnership with

What’s in it today?

• SCONE: who needs a bigger vocabulary when you can just contextualize the heck out of your n-grams?

• DAAs: Making LLMs agree with humans, one preference at a time (and sometimes only needing 5% of the data to do it!)

• CT-KL: Ignoring the KL penalty and focusing on critical tokens to boost LLMs, because sometimes you just…