MolmoAct is the first model able to “think” in three dimensions, trained efficiently and delivering benchmark-topping performance.| Ai2 Blog
How do we evaluate LLMs on underspecified queries? We show that adding clarifying context flips model rankings and uncovers model biases.| Ai2 Blog
AutoDS goes beyond standard data crunching by building upon its own findings and uncovering insights that may not be immediately apparent even to experienced researchers.| Ai2 Blog
Explore how FlexOlmo enables collaborative language model training without sacrificing data privacy or control, introducing a new, flexible approach to building shared AI models.| Ai2 Blog
Discover how SciArena is being used to evaluate foundation models’ capabilities in scientific literature tasks through community-driven, literature-grounded, and multi-disciplinary reasoning.| Ai2 Blog
Discover how OMEGA is being used to evaluate large language models' ability to generalize in math through exploratory, compositional, and transformative reasoning| Ai2 Blog
Learn how ACE is being used for seasonal forecasts and understanding decadal variations in global warming.| Ai2 Blog
We introduce a more reliable method to measure the critical batch size (CBS), analyze how CBS changes over training, and use this to train OLMo with fewer grad steps.| Ai2 Blog
Atlantes: a system of transformers for real-time GPS modeling.| Ai2 Blog
Key moments from Google Cloud Next, including our partnership with Google Cloud, OLMoTrace, and more.| Ai2 Blog
Explore the secrets of how language model developers make decisions with DataDecide.| Ai2 Blog
OLMoTrace lets you trace the outputs of language models back to their full, multi-trillion-token training data in real time.| Ai2 Blog
We announce partnership with the Cancer AI Alliance along with Google Cloud.| Ai2 Blog
Will there be a system that automatically identifies gaps in scientific knowledge and runs experiments?| Ai2 Blog
Ai2 Paper Finder is an LLM-powered literature search system that mimics the iterative paper-finding process.| Ai2 Blog
Ai2's recommendation in response to the White House’s Request for Information on an AI Action Plan.| Ai2 Blog
Ai2 is humbled to be included on Fast Company's 2025 most innovative companies list for making AI that are truly open.| Ai2 Blog
Introducing OLMo 2 32B, the most capable and largest model in the OLMo 2 family.| Ai2 Blog
Our mixture-of-experts model is available on the Apple app store! The OLMoE app allows anyone to test the model privately and securely.| Ai2 Blog
Ai2, founded by Paul Allen and led by Ali Farhadi, conducts high-impact research and engineering to tackle key problems in artificial intelligence.| allenai.org
Ai2 has been awarded a combined $152 million from the U.S. National Science Foundation (NSF) and NVIDIA as part of a jointly funded project to advance our research and develop truly open AI models and solutions that will accelerate scientific discovery.| allenai.org
A technical deep-dive into Tülu 3, with the model "recipe", data, and more.| allenai.org
Our next generation of fully-open base and instruct models sit at the Pareto frontier of performance and training efficiency. Check out our [paper](https://arxiv.org/abs/2501.00656) to learn more, or keep reading for a summary.| allenai.org
Ai2, a non-profit research institute founded by Paul Allen, is committed to breakthrough AI to solve the world’s biggest problems.| allenai.org
