▲ 0 ▼ c/gai@sopuli.xyz · cm0002@lemmy.world · 19 Jun 2025 07:49 Unpacking the bias of large language models #gai MIT News | Massachusetts Institute of Technology Unpacking the bias of large language models MIT researchers discovered the underlying cause of position bias, a phenomenon that causes large language models to overemphasize the beginning or end of a document or conversation, while neglecting the middle. They built a theoretical framework that can be used to diagnose and correct position bias in future model designs, leading to more accurate, reliable AI agents. news.mit.edu check post at origin
0 comments
log in to comment.