19 million women in the U.S. live in contraception “deserts,” meaning they do not have reasonable access to a health center that offers the full range of contraceptive methods. A team of UNC researchers has been developing an AI-driven tool known as the Sexual And Reproductive Health Assistant (SARHAchat) to encourage candid conversations about reproduction and safe sex, informed by the latest scientific evidence.

Imagine a world where AI can understand human emotions as well as a close friend. This is the vision driving Associate Professor Snigdha Chaturvedi’s research in natural language processing. Her goal is to align large language models (LLMs) more closely with human communication, making them more effective in applications where they interface with us.

Doctoral student David Wan was awarded a 2024 Google PhD Fellowship in natural language processing. The fellowship covers up to three years of tuition and fees, plus a stipend to be used for living expenses, travel, and personal equipment.

Developed by a team of Carolina researchers including Computer Science’s Shashank Srivastava, ShesPrEPared uses artificial intelligence to help women find their best option for HIV prevention.

UNC CS personnel had 14 papers accepted by the 2024 International Conference on Learning Representations. Five of the papers were selected for spotlight posters, and the submissions came from six different research groups.

Doctoral student Peter Hase joined host Sam Charrington to discuss “scalable oversight,” the importance of developing a deeper understanding of how large neural networks make decisions, how matrices are probed by interpretability researchers, and more.

UNC Computer Science researchers made a splash at the 2023 Conference and Workshop on Neural Information Processing Systems (NeurIPS). Current and future UNC CS personnel had 14 papers accepted.

In a recent paper: “Any-to-Any Generation via Composable Diffusion,” Microsoft Azure Cognitive Service Research and UNC NLP present CoDi, a novel generative model capable of processing and simultaneously generating content across multiple modalities.

UNC Computer Science research groups led by Mohit Bansal, Gedas Bertasius, and Marc Niethammer presented several papers and contributed to workshops at CVPR 2023. Here is a roundup of their papers, which used deep learning to tackle computer vision tasks.