Research Paper · 2026
Elbert Tany, Moses Handoyo, Kenneth Sean Ternadi, Anderies, Andry Chowanda
AI-Driven Code Sustainability: Architectures, Datasets, and Profiling Mechanisms for Energy-Efficient Software
A TypeScript sustainability dataset built by transpiling and profiling Python efficiency benchmarks, used to fine-tune lightweight language models with QLoRA to flag energy-inefficient code — and why the same recipe helped one model and collapsed another.
Code SustainabilityEnergy SmellsTypeScriptLarge Language Models
Read article