Pruning CXR Models
Study of how neural network pruning changes chest X-ray classification performance, model footprint, and demographic fairness.
Study of how neural network pruning changes chest X-ray classification performance, model footprint, and demographic fairness.
Agentic CLI for querying and analyzing local MLflow experiment logs with natural-language routing, Langfuse tracing, and zero-cost observability.
Course project exploring a generate-simulate-improve loop for cache-replacement policy design.
Agentic dataset discovery from natural-language requirements to ranked reports across Hugging Face and Kaggle catalogs.