New project led by Shriya Atmakuri in collaboration with @aps6992: Ai2's Asta system now reports weekly which papers its research summaries have cited.  The aim is to give credit to the work that powers the reports, and provide a dataset for studying how AI systems cite science.

Introducing Asta DataVoyager—our new AI capability in Asta that turns structured data into transparent, reproducible insights. Built for scientists, grounded in open, inspectable workflows. 🧵

A few new challengers enter SciArena—including DeepSeek-V3.2-Exp and Claude Sonnet 4.5 🔬

As part of Asta, our initiative to accelerate science with trustworthy AI agents, we built AstaBench—the first comprehensive benchmark to compare them. ⚖️

Introducing Asta—our bold initiative to accelerate science with trustworthy, capable agents, benchmarks, & developer resources that bring clarity to the landscape of scientific AI + agents. 🧵

🚀 In March, we launched Paper Finder, an LLM-powered literature search agent that surfaces papers other tools miss. Now, we’re releasing an open-source snapshot to enable others to inspect & build on it—and reproduce the results. 🧵

With fresh support of $75M from @NSF and $77M from @NVIDIA, we’re set to scale our open model ecosystem, bolster the infrastructure behind it, and fast‑track reproducible AI research to unlock the next wave of scientific discovery. 💡

🚨 SciArena update + evaluation of new models including GPT-5! 🚨 With thousands of new votes, new LLMs are reshaping our leaderboard for scientific literature tasks. o3 still leads—but GPT-5, Claude Opus 4.1, & more are closing the gap.

Are you a researcher in CS or a CS-adjacent field who could use help in refining your research ideas? Want to try our new AI-powered tool that helps with just that in a paid user study? Details and sign up here!

Great science starts with great questions. 🤔✨ Meet AutoDS—an AI that doesn’t just hunt for answers, it decides which questions are worth asking. 🧵

We’ve upgraded ScholarQA, our agent that helps researchers conduct literature reviews efficiently by providing detailed answers. Now, when ScholarQA cites a source, it won’t just tell you which paper it came from–you’ll see the exact quote, highlighted in the original PDF. 🧵

This was a fun collaboration led by @YilunZhao_NLP and @kaiyan_z from @armancohan's lab at Yale University. Annotators preferred o3 in our study, which was found to give more detailed and technical answers. Curious to see if community voting changes the picture!

Introducing SciArena, a platform for benchmarking models across scientific literature tasks. Inspired by Chatbot Arena, SciArena applies a crowdsourced LLM evaluation approach to the scientific domain. 🧵

@allen_ai @SemanticScholar is hiring an #ml #nlp #ai reasoning researcher for a Research Scientist, Agents for Science position with target start dates in 2025. Excited about developing AI systems with deep reasoning capabilities for science? Send an application our way!

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. DataDecide opens up the process: 1,050 models, 30k checkpoints, 25 datasets & 10 benchmarks 🧵

For years it’s been an open question — how much is a language model learning and synthesizing information, and how much is it just memorizing and reciting? Introducing OLMoTrace, a new feature in the Ai2 Playground that begins to shed some light. 🔦

Imagine AI doing science: reading papers, generating ideas, designing and running experiments, analyzing results… How many more discoveries can we reveal? 🧐 Meet CodeScientist, a promising next step toward autonomous scientific discovery. 🧵

Meet Ai2 Paper Finder, an LLM-powered literature search system. Searching for relevant work is a multi-step process that requires iteration. Paper Finder mimics this workflow — and helps researchers find more papers than ever 🔍

Announcing OLMo 2 32B: the first fully open model to beat GPT 3.5 & GPT-4o mini on a suite of popular, multi-skill benchmarks. Comparable to best open-weight models, but a fraction of training compute. When you have a good recipe, ✨ magical things happen when you scale it up!