Dissertation Defenses

Algorithm-Hardware Co-design For Efficiency Optimization Of Machine Learning Workloads

Increasing The Practicality Of Verification Using Incomplete Solutions

Streaming Sketching: Mathematical Theory and Practical Algorithms

Tool-Use Robot Manipulation Tasks for Cooperative and Explainable Operations in Safety-Critical Domains

Data-Aware Computing for Efficient Extended Reality Systems

Learning Dense Visual Features for the Sun and Natural Scenes

Exploring the Effects of Memory Vulnerabilities Across the Computer Architecture Stack

Practical Verification of Distributed Systems: Streamlining Safety Proofs Using Invariant Taxonomies, and Verifying Latency Properties Using Symbolic Latency

Enhancing Video Understanding Through Deep Generative Models and Task Comprehension

Leveraging Compositional Structure for Reinforcement Learning and Decision Making Problems

Towards Enhanced Human-AI Interaction: A Holistic Approach to Personalization in Natural Language Processing

Assisting Cloud System Development with Automated Insight Generation

Exploring Intersectionality in Responsible Data Management

Improving Graph Representation Learning with Augmentations, Uncertainty Quantification and Large Language Model Guidance

Exploiting Tree Structure in Empirical Game-Theoretic Analysis for Extensive-Form Games

Bridging Data and Hardware Gap for Efficient Machine Model Scaling

Novel Sensing Approach for Shoulder Kinematics – Exploring Kirigami-Inspired Sensing and Machine Learning Platforms for At-Home Joint Monitoring

Coherent Physical Commonsense Reasoning in Foundational Language Models

Combating Randomness: Towards Efficient Data-Intensive Applications Using Software-Hardware Co-Design

Reconstructing Full 3D Geometry and Interaction Dynamics from Real World data

Learning to Interact with the 3D World

Three Lenses on Improving Programmer Productivity: From Anecdote to Evidence

Enabling Robot Arm Motion with Multiple Objectives Through Agent Representation and Reasoning

Quality of Service for Performance-Critical Cloud Applications

Interpretability, Controllability, and Expressivity in the Age of Language Models