SOLAR Lab Seminar: Aditya Akella, Man-Made Heuristics Are Dead. Long Live Code Generators!

Content

Image

Aditya Akella

Speaker

Aditya Akella (UT Austin)

Description

This work suggests a focus on AI’s ability to generate code and replace traditionally human-designed components (heuristics) in research. This connects to the seminar by showcasing how AI is moving beyond simple augmentation to automating core tasks within experimental design and execution, forcing researchers to redefine the role of human expertise.

Abstract

Policy design for various systems controllers has conventionally been a manual process, with domain experts carefully tailoring heuristics for the specific instance in which the policy will be deployed. 

In this paper, we re-imagine policy design via a novel automated search technique fueled by recent advances in generative models, specifically Large Language Model (LLM)-driven code generation. We outline the design and implementation of PolicySmith, a framework that applies LLMs to synthesize instance-optimal heuristics. We apply PolicySmith to two long-standing systems policies - web caching and congestion control, highlighting the opportunities unraveled by this LLM-driven heuristic search. 

For caching, PolicySmith discovers heuristics that outperform established baselines on standard open-source traces. For congestion control, we show that PolicySmith can generate safe policies that integrate directly into the Linux kernel.

Paper