AI Briefing#ai2
The latest AI and developer news about #ai2, with the original source and a short summary.
The latest AI and developer news about #ai2, with the original source and a short summary.
The benchmark separates skill learning from familiar setup success, revealing that current systems fail when block positions or stack heights change.
AI2 introduces StateBench, a four-level benchmark revealing that while recurrent architectures generalize better than transformers, top models like Gemini 3.8 Flash and Claude Opus 5.5 still fail on hard state-tracking tasks.
In a study with 17 researchers, domain experts rated ideas developed with LitPivot as better grounded in literature than those developed with a text editor and AI chat tool.
The model generates 128-dimensional vectors for every 10-meter pixel on Earth, outperforming AlphaEarth Foundations and Tessera v2 across six continents.
The 13-million-case benchmark tests 15 models and finds that existing debiasing techniques often fail to fix this identity-based vulnerability without sacrificing accuracy.
The new system generates global 3 km climate data at approximately 1/300th the cost of the X-SHiELD atmospheric model.
Introducing GPU time budgets and fair-share allocation instead of priorities significantly improves wait times while maintaining 98% cluster utilization
The new system reduced p90 debug workload wait times from 2 hours to 30 seconds and cut on-call maintenance toil by 74% across clusters of thousands of NVIDIA GPUs.
The 8B model, based on Qwen3-8B, generates reports in an average of 51.1 seconds compared to 178.5 seconds for the previous Claude-based Thinking mode.
Ai2 releases Olmo-core 3, an open-source training framework designed for large MoE models, achieving 2.7x throughput improvement in preliminary tests and benchmarked at over 1 trillion parameters.
AI2 has released BenchMIRT, a multidimensional IRT-based tool that analyzes which specific capabilities individual questions in LLM benchmarks actually measure.
Ai2 has unveiled a geospatial foundation model and inference platform that processes large-scale satellite data.
AI2 has launched olmo-eval, a new workbench designed to streamline repetitive evaluation during the model development process.
Ai2 released EMO, showing that performance can be largely preserved using only a subset of a MoE's experts.
AI2 released MolmoAct 2, an open foundation model for real-world robots, along with a dataset of over 720 hours.
ScienceWorld and DiscoveryWorld reveal how AI science agents actually perform.
AI2 unveiled MolmoWeb, an open web agent that supports everything from data to deployment.