Seminar: Sanjay Shakkottai, Towards Discrete Diffusion Models for Language and Image Generation

Content

Speaker 

Sanjay Shakkottai (University of Texas at Austin)

Abstract

We discuss discrete diffusion models that offer a unified framework for jointly modeling categorical data such as text and images. We present a new model that we have developed for language generation called the Anchored Diffusion Language Model (ADLM). ADLM is grounded in a novel two-stage framework that first predicts distributions over important tokens via an anchor network (e.g., key words or low-frequency words that anchor a sentence), and then predicts the likelihoods of missing tokens conditioned on the anchored predictions. ADLM significantly improves test perplexity on LM1B and OpenWebText, achieving up to 25.4% gains over prior DLMs, and narrows the gap with strong AR baselines. It also achieves state-of-the-art performance in zero-shot generalization across seven benchmarks and surpasses AR models in MAUVE score, which marks the first time a DLM generates better human-like text than an AR model. Beyond diffusion, anchoring boosts performance in AR models and enhances reasoning in math and logic tasks, outperforming existing chain-of-thought approaches. Project page: https://anchored-diffusion-llm.github.io/  

Speaker Bio

Sanjay Shakkottai received his PhD from the ECE Department at the University of Illinois at Urbana-Champaign in 2002. He currently works at The University of Texas at Austin, where he is a Professor in the ECE and CS Departments, and holds the Cockrell Family Chair in Engineering #15. He is also the Director of the Center for Generative AI, a campus-wide computing cluster at UT Austin. He received the NSF CAREER award in 2004 and was elected as an IEEE Fellow in 2014. He was a co-recipient of the IEEE Communications Society William R. Bennett Prize in 2021. He has served as the Editor in Chief of IEEE/ACM Transactions on Networking. His current research interests are in diffusion models and Generative AI, with applications in language models, image editing, and decision-making in wireless networks.