Ari Holtzman

Ari Holtzman is an Assistant Professor of Computer Science at the University of Chicago. His research spans dialogue systems, text generation, and foundational AI methodologies, including the development of Nucleus Sampling and contributions to the Amazon Alexa Prize. He holds an interdisciplinary degree from NYU in Computer Science and Philosophy of Language, and is nearing completion of his PhD at the University of Washington.

Research interests include generative models, alignment challenges in LLMs, evaluation metrics like CLIPScore, and model efficiency techniques such as Qlora finetuning. His work bridges theoretical insights with practical applications, emphasizing both technical innovation and ethical considerations in AI.

Key awards include the 2017 Amazon Alexa Prize and Phi Beta Kappa honors at NYU. His recent publications focus on benchmarking frameworks, cache optimization for large models, and understanding model limitations through AbsenceBench. Research contributions extend to multimodal systems, computational creativity, and machine unlearning protocols.