Kiera Charlotte Lai
Computer systems engineering
Hometown: Pleasanton, California, United States
Graduation date: Spring 2028
FURI | Spring 2025
Optimizing Large Language Models to Minimize Attention Latency for Increasingly Complex Inputs
This project aims to improve Large Language Models’ context retention and coherence by optimizing attention mechanisms, altering parameters, and amending inputs. Unlike previous studies, this approach focuses on modifying context windows and word graph parameters to augment the model’s processing of inputs beyond restricted context retention. A simplified version of an AI model will be designed and evaluated for improved attention based on relevance and output changes. By enhancing AI attention span, this research aims to advance LLMs’ contextual awareness, ameliorating the conversational style of generative AI. These findings could contribute to the broader advancement and reliability of AI applications.
Mentor: Adwith Malpe, Ryan Meuth
It’s hip to be square.
Students presenting projects at the Fulton Forge Student Research Expo are encouraged to download this personal QR code and include it within your poster. This allows expo attendees to explore more about your project and about you in the future.
Right click the image to save it to your computer.