HN
login
Do small language models know what they don't know?arxiv.org

We explore whether entropy-based confidence signals can be leveraged to improve the accuracy of Small Language Models (SLMs) with fewer than 3 billion parameters, running entirely on consumer hardware. We evaluate…

3 points•Brajeshwar•8 days ago•0 comments•
#language models#small language#small language models#models know#language models know#whether#entropy#explore

0 comments

No comments yet.

Related stories