Practice Mode
"Practice mode" in machine learning and related fields encompasses the development and application of methods to improve model performance and robustness through iterative refinement and adaptation. Current research focuses on bridging the gap between theoretical guarantees and practical performance, exploring techniques like continual learning, transfer learning, and reinforcement learning with various model architectures (e.g., neural networks, graph neural networks, large language models). This research is significant because it addresses the limitations of traditional training paradigms, leading to more efficient, adaptable, and reliable AI systems across diverse applications, from language processing and image recognition to robotics and personalized education.
Papers
LLM Pruning and Distillation in Practice: The Minitron Approach
Sharath Turuvekere Sreenivas, Saurav Muralidharan, Raviraj Joshi, Marcin Chochowski, Mostofa Patwary, Mohammad Shoeybi, Bryan Catanzaro, Jan Kautz, Pavlo Molchanov
Graph Classification via Reference Distribution Learning: Theory and Practice
Zixiao Wang, Jicong Fan
Operationalizing the Blueprint for an AI Bill of Rights: Recommendations for Practitioners, Researchers, and Policy Makers
Alex Oesterling, Usha Bhalla, Suresh Venkatasubramanian, Himabindu Lakkaraju
Investigating Public Fine-Tuning Datasets: A Complex Review of Current Practices from a Construction Perspective
Runyuan Ma, Wei Li, Fukai Shang
A Systematic Review of Generative AI for Teaching and Learning Practice
Bayode Ogunleye, Kudirat Ibilola Zakariyyah, Oluwaseun Ajao, Olakunle Olayinka, Hemlata Sharma
A Survey on Compositional Learning of AI Models: Theoretical and Experimetnal Practices
Sania Sinha, Tanawan Premsri, Parisa Kordjamshidi