https://ai-news.ypetit.net/sources/ai2/ Latest publications from Allen Institute for AI (Ai2), aggregated by AI News from the official source. en Fri, 23 Oct 2026 19:08:30 GMT https://allenai.org/blog/litpivothttps://allenai.org/blog/litpivotFri, 23 Oct 2026 08:00:00 GMTresearchLitPivot helps researchers refine early research ideas through literature-supported critiques and revisions as their proposals evolve. https://allenai.org/blog/olmoearth-v1-3https://allenai.org/blog/olmoearth-v1-3Thu, 22 Oct 2026 08:00:00 GMTresearchOlmoEarth v1.3 brings state-of-the-art satellite embeddings, precomputed global data, and a browser-based Embeddings Explorer for mapping and analyzing the planet. https://allenai.org/blog/planteurhttps://allenai.org/blog/planteurThu, 22 Oct 2026 08:00:00 GMTresearchPLANTEUR extends the PASTIS crop-mapping benchmark to six tropical regions to help evaluate geospatial models across more diverse agricultural systems. https://allenai.org/blog/fairstresshttps://allenai.org/blog/fairstressTue, 20 Oct 2026 08:00:00 GMTresearchFairStress is a new benchmark that reveals demographic differences in how LLMs defend and explain their decisions, even when their final choices appear fair. https://allenai.org/blog/hiro-acev2https://allenai.org/blog/hiro-acev2Wed, 14 Oct 2026 08:00:00 GMTresearchHiRO-ACEv2 captures extreme rain, wind, and pressure events at 3 km resolution for about 1/300th the cost of X-SHiELD, making detailed global climate simulations more accessible to researchers. https://allenai.org/blog/molmo2fishhttps://allenai.org/blog/molmo2fishTue, 13 Oct 2026 08:00:00 GMTresearchMolmo 2’s open release helped MIT researcher Kai Van Brunt develop Molmo2Fish, an experimental model that tracks salmon in sonar footage and corrects some tracking errors using everyday language. https://allenai.org/blog/impactful-schedulinghttps://allenai.org/blog/impactful-schedulingFri, 09 Oct 2026 08:00:00 GMTresearchWe explain how Ai2’s new GPU scheduler uses time budgets, fair-share allocation, and time-slicing to prioritize high-impact research, shorten queue waits, and keep GPUs busy. https://allenai.org/blog/colm-2026-recaphttps://allenai.org/blog/colm-2026-recapFri, 09 Oct 2026 08:00:00 GMTresearchAi2 joined COLM 2026 to share advances in open language models, training infrastructure, and AI for science, including Olmo Hybrid, Asta, and Bolmo, while connecting with the broader research community. https://allenai.org/blog/bolmo-naturehttps://allenai.org/blog/bolmo-natureWed, 07 Oct 2026 08:00:00 GMTresearchThe technique behind Bolmo, Ai2’s fully open byte-level language models, is now published in Nature, with new checkpoints showing the approach generalizes beyond Olmo to other model families. https://allenai.org/blog/astabriefhttps://allenai.org/blog/astabriefFri, 02 Oct 2026 08:00:00 GMTresearchWe’re releasing AstaBrief, an 8B open-weights model for generating cited scientific reports, available in Asta’s Fast mode or to download and run on your own infrastructure. https://allenai.org/blog/olmocore3https://allenai.org/blog/olmocore3Thu, 01 Oct 2026 08:00:00 GMTresearchOlmo-core 3 introduces a redesigned, fully open training stack for efficiently scaling mixture-of-experts models into the trillion-parameter range. https://allenai.org/blog/olmo-arenahttps://allenai.org/blog/olmo-arenaThu, 17 Sep 2026 08:00:00 GMTresearchA crowdsourced game built on Olmo 3 showed how people can exploit unexpected model behaviors to stress-test prosocial AI evaluations—and how open access to a model’s internals can help researchers understand why those tests break. https://allenai.org/blog/autodiscovery-student-challengehttps://allenai.org/blog/autodiscovery-student-challengeMon, 14 Sep 2026 08:00:00 GMTresearchUniversity of Washington students put Ai2’s AutoDiscovery to the test, showing how AI can surface promising scientific leads while making human judgment, domain expertise, and rigorous validation more important than ever. https://allenai.org/blog/goodfire-olmohttps://allenai.org/blog/goodfire-olmoWed, 09 Sep 2026 08:00:00 GMTresearchGoodfire used Ai2’s fully open post-training stack to predict LLM behavioral changes, trace unwanted model behavior back to individual training examples, and test targeted fixes without sacrificing broader capability gains. https://allenai.org/blog/benchmirthttps://allenai.org/blog/benchmirtTue, 01 Sep 2026 08:00:00 GMTresearchBenchMIRT is a new method for auditing LLM benchmarks question by question, revealing which capabilities they actually measure and helping researchers build smaller, more focused, and easier-to-interpret evaluations. https://allenai.org/blog/swedish-autodiscovery-recaphttps://allenai.org/blog/swedish-autodiscovery-recapTue, 01 Sep 2026 08:00:00 GMTresearchAt an Ai2 event marking our expanded collaboration with Providence Swedish, researchers explored the hardest problems in AI-assisted science: keeping systems steerable, grounded in human judgment and sound methods, and responsive to new evidence and experiments. https://allenai.org/blog/swedish-autodiscovery-partnershiphttps://allenai.org/blog/swedish-autodiscovery-partnershipThu, 27 Aug 2026 08:00:00 GMTresearchAi2 and Providence Swedish Cancer Institute are expanding their collaboration after AutoDiscovery helped researchers uncover and validate a promising new immune signal in invasive lobular breast cancer. https://allenai.org/blog/thai-llm-dolmahttps://allenai.org/blog/thai-llm-dolmaWed, 26 Aug 2026 08:00:00 GMTresearchThai researchers adapted Ai2’s open Dolma toolkit to build Mangosteen, a 47-billion-token Thai corpus that filters low-quality web data while maintaining or improving model performance and strengthening Thai cultural knowledge.