New Framework
Recent research focuses on developing versatile frameworks for various tasks, primarily aiming to improve efficiency, reproducibility, and accessibility within their respective domains. These frameworks leverage diverse techniques, including programmatic data generation for LLMs, deep learning architectures for image and audio processing, and reinforcement learning for optimization and automated testing. The resulting advancements enhance the development and evaluation of AI models, improve the reliability of benchmarking processes, and offer new tools for diverse applications ranging from healthcare diagnostics to autonomous vehicle navigation.
Papers
ARLO: A Framework for Automated Reinforcement Learning
Marco Mussi, Davide Lombarda, Alberto Maria Metelli, Francesco Trovò, Marcello Restelli
SALTED: A Framework for SAlient Long-Tail Translation Error Detection
Vikas Raunak, Matt Post, Arul Menezes
CertiFair: A Framework for Certified Global Fairness of Neural Networks
Haitham Khedr, Yasser Shoukry