Introduction
Instruction tuning has enabled modern foundation models to function beyond next-word prediction to perform a wide range of downstream tasks. This advancement has led to the creation of impactful industrial applications such as ChatGPT, Claude, and Gemini. It has also spurred increased discussion within the open-source and research communities, resulting in the creation of new benchmarks and resources, the development of new training methods, and efforts to understand the limitations of these methods.
In the latest trend, we are witnessing the remarkable effectiveness of instruction-following extending beyond language only, to multi-modal settings, including advancement in applications such as image editing and robotic command execution.
To facilitate discussions on advancing instruction tuning methodologies and constructing general-purpose instruction-following models, we propose to organize the second workshop on Instruction Tuning and Instruction Following at NeurIPS 2024. It is crucial to organize this workshop due to the prevalence of proprietary models with restricted access and limited transparency regarding their training processes. This creates a pressing need for an open platform to encourage discussions and collaborations.
Confirmed Speakers
- Chelsea Finn, Stanford University, Physical Intelligence
- Danqi Chen, Princeton University
- Eric Wallace, OpenAI
- Manling Li, Stanford University, Northwestern University
- Minjoon Seo, Korea Advanced Institute of Science and Technology, Twelve Labs
- Omer Levy, Tel Aviv University, Meta AI
- Hannah Rose Kirk, University of Oxford
Confirmed Reviewers
- Lj Miranda, Allen Institute for AI
- Bowen Zhao, University of Washington
- Hamish Ivison, Allen Institute for AI
- Shane Lyu, Allen Institute for AI
- Mickel Liu, Allen Institute for AI
- Harvey Yiyun Fu, University of Southern California
- Ting-Yun Chang, University of Southern California
- Mozhdeh Gheini, University of Southern California
- Huihan Li, University of Southern California
- Lorena Yan, University of Southern California
- Siyuan Wang, University of Southern California
- Xiaoyue Xu, Tsinghua University
- Zhankui He, Google
- Bowen Pan, Massachusetts Institute of Technology
- Taiwei Shi, University of Southern California
- Xingyao Wang, University of Illinois Urbana-Champaign
- Yu Wang, University of California, San Diego
- Seonghyeon Ye, Korea Advanced Institute of Science and Technology
- Lintang Sutawika, Eleuther AI & Carnegie Mellon University
- Sadhika Malladi, Princeton University
- Tianyu Gao, Princeton University
- Zhiyuan Zeng, Tsinghua University
- Dingli Yu, Princeton University
- Simran Kaur, Princeton University
- Christopher Klamm, University of Mannheim
- Will Brannon, Massachusetts Institute of Technology
- Da Yin, University of California, Los Angeles
- Jinheon Baek, Korea Advanced Institute of Science and Technology
- Sungdong Kim, Korea Advanced Institute of Science and Technology & NAVER Cloud
- Akari Asai, University of Washington
Organizers
- Mengzhou Xia, Princeton University
- Zexue He, University of California, San Diego
- Marzieh Fadaee, Cohere For AI
- Seungone Kim, Korea Advanced Institute of Science and Technology, Carnegie Mellon University
- Qinyuan Ye, University of Southern California
- Yizhong Wang, University of Washington
- Shayne Longpre, Massachusetts Institute of Technology
- Daniel Khashabi, Johns Hopkins University