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 · TEC 340
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.
#ComputerVision#VisionLanguageModels#RAG#ConversationalMemory#AIAgents#TrustworthyAI#HumanCenteredAI
The scoring target a memory benchmark picks can flip which LLM wins, even when the overall ranking looks stable.
Which Gaussians you pick as anchors barely matters: cheap samplers match the default strategy at deployment budgets, so the standard 8,192-anchor choice over-spends compute.
Plain-language write-ups of the lab's papers are on the Writing page.