Hi! I'm a predoctoral researcher on the AllenNLP team at Ai2, where I work broadly on problems in machine learning, natural language processing, and AI policy. I've been awarded an NSF Computer Science Graduate Fellowship, and I'm advised by Pradeep Dasigi and Jesse Dodge.

Research Interests

My research generally spans machine learning and natural language processing. I work to improve and better understand modern language models, and I've recently focused on efficiently adapting them to specialized domains, building better evaluations, and measuring and reporting the environmental impact of AI.

I'm also a strong supporter of open science. I've contributed to various openly released tools and platforms including OLMo, Dolma, RewardBench, and OLMoE, with more coming soon.

Public Policy

I spend a portion of my time helping policymakers understand and address the societal impacts of advances in AI. I started and currently lead Ai2's public policy efforts, through which I regularly engage with policymakers at the local, state, and federal levels. I previously served on the City of Seattle's Generative AI Policy Advisory Group.

Education & Experience

I received my masters in computational linguistics and bachelors in computer science from the University of Washington, where I worked with Jungo Kasai and Yangfeng Ji, and was advised by Noah Smith.

I've previously worked on code & program synthesis at X, the moonshot factory and Google Labs, language + vision models at Ai2, and platform health at Twitter. I also spent a few years as a software engineer at Tableau and Google. See my CV for more details.

Service

In my free time, I serve in a number of positions in the UW Alumni Association including as a District Dawg, on the GOLD Council, and as an early career mentor. I'm also on the City University Community Advisory Committee, where I help advise on major capital projects on UW's campus as well as in the greater University District.

Latest News

September 2024 Our paper on unsettled law in the age of generative AI was accepted to the Journal of Law and Technology at Texas!
September 2024 We're releasing OLMoE, a state of the art open mixture-of-experts language model. Check it out here!
August 2024 Our paper on the first amendment rights of large language models was accepted to the University of North Carolina's First Amendment Law Review!
August 2024 OLMo won a Theme Paper award and Dolma won a Best Resource Paper award at ACL 2024!
May 2024 OLMo and Dolma were both accepted to ACL 2024!
March 2024 We released RewardBench, the first holistic evaluation suite for reward models. Read more here, and check out the leaderboard and repo!
February 2024 We released OLMo 7B and Dolma, more coming soon! Read more about both here.
November 2023 I co-presented on responsible, real world AI development and deployment to the Washington State AI Community of Practice
November 2023 The City of Seattle released their Generative Artificial Intelligence Policy, which I contributed to through the City's Generative AI Policy Advisory Group.
August 2023 I was awarded a National Science Foundation Computer Science Graduate Fellowship, equivalent in funding to the NSF GRFP.
August 2023 I helped organize a panel discussion on the future of AI development and regulation with the office of Senator Maria Cantwell.

If you're looking for Jacob Morrison the musician, his website is located at https://jacobmorrison.me/.