Same Ranking, Different Winner
The scoring target a memory benchmark picks can flip which LLM wins, even when the overall ranking looks stable.
Artificial Intelligence & Multimodal Systems
Lab @ USM · CSCE
A research group at the University of Southern Mississippi building AI systems that learn from images, video, text, speech, and sensor data, and stay reliable under uncertainty.
The scoring target a memory benchmark picks can flip which LLM wins, even when the overall ranking looks stable.
A fixed compression layer in a RAG pipeline collapses the reader-scaling trend that the uncompressed setup shows.
Recovery affordances that let an agent notice and repair tool calls that failed silently.
A cross-domain survey of power-line, crack, and retinal-vessel segmentation, from classical pipelines to promptable foundation models.
Measuring how much typed user speech gets absorbed into edited artifacts at unmarked within-turn seams.