The Columbia Center of AI Technology in collaboration with Amazon is pleased to announce its 2024 Call for Research Project Proposals and Fellowship Nominations.
Founded in 2020, the Center is a collaborative effort between Columbia and Amazon with a mission to better society through the development and adoption of advanced AI technology contributing to a more secure, connected, creative, sustainable, healthy and equitable humanity. This year, the center will fund multiple research projects and new Doctoral Fellows through generous support from Amazon.
In the 2024 cycle, the Center will award projects and fellowships in the following research focus areas:
- Reducing hallucination in LLMs
- Improving holistic evaluation of text and image generation models
- Responsible generative AI
- This may include, but is not limited to, hallucination, measurement and mitigation, guardrail models, privacy and IP concerns, and machine unlearning and model disgorgement.
- Novel model architectures for improved performance (accuracy & efficiency)
- Improved distillation approaches to create performant small models
- Foundations of AI and Machine Learning
- Synthetic data generation
Research Project Proposals
Please see the InfoReady page for more information about topics of interest, proposal eligibility, and application instructions. Proposals are due via InfoReady by February 29th, 2024.
Pre-proposal Q&A: Want to learn more about the call and discuss your ideas with an Amazon researcher? Columbia faculty are invited to join us for the virtual Q&A session with Amazon scientists on Monday, February 12th at 3pm.
A list of past Funded Research Projects can be found on the Funded Projects page.
PhD Fellowships
Please see the Call for Nominations InfoReady page for more information about topics of interest, student eligibility, and nomination instructions. Nominations are due by February 29th, 2024.
Faculty with questions about the proposal or nomination process are encouraged to email Jessie Taft, CAIT Program Manager, at [email protected]