Ai2 Newsletter

To effectively conserve and sustainably manage our critical ocean resource, we must maintain a watchful eye on maritime activity. We're open-sourcing Atlantes, a powerful new suite of AI models that use GPS data to analyze vessel behavior in real time. We turned massive, unstructured, noisy, and irregular GPS data into real-time information to help understand vessel behaviors like fishing, transiting, and anchoring.

This model is one component in a complex and global effort to monitor the planet's oceans for sustainability and conservation. We hope that by open-sourcing Atlantes, we can enable other researchers and environmentalists to better understand the strengths and limitations of this approach and also foster more widespread adoption of maritime transparency.

Ever wonder how LLM developers choose their pretraining data? It’s not guesswork— all AI labs create small-scale models as experiments, but the models and their data are rarely shared.

To empower open exploration of these questions, we released DataDecide, a suite of 1,050 models, 30k checkpoints, 25 datasets, and 10 benchmarks. We evaluate all models across a suite of 10 downstream tasks and calculate how accurately we can use small models to predict that one pretraining corpus will lead to better performance than another for our largest models.

Which benchmarks should we be evaluating on? Are scaling laws better? Can we use a better metric than accuracy? Read the blog to see our recommendations for model developers.

"With OLMoTrace, we’re actually bringing accessibility to openness, enabling everybody to start looking into the inner workings of the relationships between the input and output of these models." – Ali Farhadi, Ai2 CEO