Mentor Areas
Distributed systems
Computer networks
Systems for AI
Description:
AI-driven System Optimizers, such as AlphaEvolve and GEPA, enable the autonomous generation and optimization of complex algorithms and systems. However, existing frameworks are constrained by limited scalability and computational inefficiencies, often struggling to adapt to high-dimensional search spaces. The project focuses on two layers. At the application layer, we will design techniques for system evolution that leverage parallelism and collective information to significantly reduce resource usage compared to traditional methods. At the model layer, we plan to develop adaptive transformer architectures, specifically Speculative Decoding and Early Exit, to optimize computational efficiency.
This project builds on top of OpenEvolve and vLLM.
Preferred Qualifications
Extensive programming experience (particularly in Python)
Experience with building systems and optimization for performance and scalability
Experience with OpenEvolve and/or vLLM.
Details:
Preferred Student Year
Second-Year, Junior, Senior
Academic Term
Fall, Spring, Summer
I prefer to have students start during the above term(s).Volunteer
Yes
Yes indicates that faculty are open to volunteers.Paid
No
Yes indicates that faculty are open to paying students they engage in their research, regardless of their work-study eligibility.Work Study
No
Yes indicates that faculty are open to hiring work-study-eligible students.