Ai2 blog
November 2024 - OLMo 2: The best fully open language model to date
Our next generation of fully-open base and instruct models sit at the Pareto frontier of performance and training efficiency.
November 2024 - Tülu 3 opens language model post-training up to more tasks and more people
Tülu 3 is a leading instruction following model family, offering fully open-source data, code, and recipes designed to serve as a comprehensive guide for modern post-training techniques.
November 2024 - Tülu 3: The next era in open post-training
A technical deep-dive into Tülu 3, with the model "recipe", data, and more.
Chris Bretherton / December 2024 - The Ai2 Climate Emulator
The Ai2 Climate Emulator: Fast, accurate AI-based atmospheric simulation for historical and future climates.
Akari Asai / November 2024 - Ai2 OpenScholar: Scientific literature synthesis with retrieval-augmented language models
Ai2’s & UW’s OpenScholar, a retrieval-augmented LM, helps scientists navigate and synthesize scientific literature.
Sahil Verma, Yanai Elazar / November 2024 - How many Van Goghs does it take to Van Gogh? Finding the imitation threshold
Meet MIMETIC^2: Finding the number of images required by text-to-image models for imitation of a concept.
Lj Miranda / October 2024 - Hybrid preferences: Learning to route instances for human vs. AI feedback
We introduce a routing framework that combines inputs from humans and LMs to achieve better annotation quality.
Yuling Gu / October 2024 - Applying theory of mind: Can AI understand and predict human behavior?
"Theory of Mind" is the ability to understand that others have their own thoughts and beliefs.
Jordan Steward / October 2024 - Ai2 at COP 16: Harnessing AI and conservation tech to protect our planet
We're heading to UN Biodiversity COP to showcase how open, collaborative AI can galvanize communities.
Will Merrill / October 2024 - Investigating pretraining dynamics and stability with OLMo checkpoints
We use data from our open pretraining runs to test hypotheses about training dynamics in OLMo checkpoints.
Niklas Muennighoff / September 2024 - OLMoE: An open, small, and state-of-the-art mixture-of-experts model
Introducing OLMoE, the first model to be on the Pareto frontier of performance and size, released with open data.
Yuling Gu / August 2024 - Digital Socrates: Evaluating LLMs through explanation critiques
Digital Socrates is an evaluation tool that can characterize LLMs' explanation capabilities.