I work at the boundary where silicon meets intelligence—from the circuits that power computation to the algorithms that learn from it. My interests span the full stack: gates, pipelines, instructions, and the models they enable.
-
Learning Systems
- Investigating how networks encode and compress information
- Building attention mechanisms and sequence models
- Exploring the geometry of learned representations
-
Performance & Execution
- Modeling instruction flow and speculation behavior
- Analyzing micro-architectural bottlenecks
- Optimization strategies across the compute stack
-
Circuit Design
- Layout and simulation of custom logic
- Physical implementation workflows
- Transistor-level design exploration
Interested in collaboration on performance modeling, learning architectures, or hardware-software co-design projects.
