Artifact for "GrowthHacker: Automated Off-Policy Evaluation Optimization Using Code-Modifying LLM Agents"

Description

This artifact contains the complete replication package for "GrowthHacker: Automated Off-Policy Evaluation Optimization Using Code-Modifying LLM Agents." It provides all necessary code, data, pre-computed models, and documentation to fully reproduce the experimental results presented in the paper.

Purpose

This replication artifact enables researchers to:

• Reproduce all 504 experimental runs reported in the paper
• Verify the performance metrics across different frameworks (default, AutoGen, CrewAI, two_agent)
• Extend the benchmarking to additional notebooks or LLM configurations
• Build upon the GrowthHacker system for future research

The artifact addresses the reproducibility requirements for evaluating automated code optimization in Off-Policy Evaluation scenarios. Please see more details in README.md.