DAF-MIT AI Accelerator Research Projects Featured at IEEE HPEC Conference

The IEEE High Performance Extreme Computing (HPEC) free virtual conference (Sep 25-29) features many world-class Department of the Air Force-MIT AI Accelerator research projects in a 125+ presentation agenda with 30+ talks on AI. Register now at IEEE-HPEC.org.

DAF HPEC Highlights include:
Keynote Talk: Eileen Vidrine (Air Force Chief Data & AI Officer)
Invited Talk: Lt. Col. Dr. Sean Atkins (USAF) – Integration of Effort in National Cyber Defense

AIA Sponsored Research Project: Fast AI (PI: Prof. Leiserson)
• Special Session – Large Language Models
An Analysis of Energy Requirements for Computer Vision Algorithms [Outstanding Student Paper Award]
From Words to Watts: Benchmarking the Energy Costs of Large Language Model Inference
Optimizing Compression Schemes for Parallel Sparse Tensor Algebra

AIA Sponsored Research Project: Better Networks (PI: Prof. Pentland)
• Special Session – BRAINS: Boosting Resilience through Artificial Intelligence for Networked Systems
Deployment of Real-Time Network Traffic Analysis using GraphBLAS Hypersparse Matrices and D4M Associative Arrays [Outstanding Paper Award]
Hypersparse Traffic Matrix Construction using GraphBLAS on a DPU [Outstanding Short Paper Award]
Focusing and Calibration of Large Scale Network Sensors using GraphBLAS Anonymized Hyperspace Matrices
Mapping of Internet “Coastlines” via Large Scale Anonymized Network Source Correlations

AIA Sponsored Research Project: Magnetic Navigation (PI: Prof. Edelman)
Continuous Deep Equilibrium Models: Training Neural ODEs Faster by Integrating Them to Infinity [Outstanding Student Paper Award]

AIA Sponsored Research Project: Objective Performance Prediction (PI: Prof. Heldt)
Scalable Deep Learning for Pilot Performance Analysis Using Multimodal Physiological Time Series

AIA Sponsored Research Project: AI Education (PI: Prof. Breazeal)
• Special Session – Scaling HPC Education

AIA Sponsored Research Projects: Earth Intelligence Engine (PI: Prof. Newman) & Few-Shot and Continual Learning (PI: Prof. Soljacic)
Meta-Learning and Self-Supervised Pretraining for Storm Event Imagery Translation
Robust Fine-Tuning of Vision-Language Models for Domain Generalization
Contextualizing Enhances Gradient Based Meta Learning for Few Shot Image Classification

AIA Phantoms
• Lt. Taylor Hilliard, Capt. Rebekah Magness, Captain, Maj. Jordan Tribble, Major (USAF) – Human Capital Risk Frameworks
• Braden Eichmeier (USAF) – Big Data Opportunities in Production Records at Air Force Maintenance Depots