The Summer of AI Research 2026 is now accepting applications! In this online event (July 13 - August 16) we invite people with little research experience to contribute to open source AI research under the mentorship of experienced researchers. www.eleuther.ai/soar

Looking for EleutherAI @iclr-conf.bsky.social? Come by our posters! If you're in our discord, we have a thread #general > ICLR 2026 Meetup you can join to coordinate with @stellaathena.bsky.social, Goncalo Paulo, @norabelrose.bsky.social, and members of our community who will be there!

We at @cdt.org have an amazing lineup for our 3/26 event on AI and internet scraping with @gtowntechlaw.bsky.social, including @archive.org, @wikimediafoundation.org, @cloudflare.social, @sparcopen.bsky.social, @nytimes.com, @eleutherai.bsky.social, and more! RSVP: docs.google.com/forms/d/e/1F...

Kevin Bankston@bankston.bsky.social · 5mo ago

Internet + AI policy friends: RSVP to join @cdt.org at @georgetownlaw.bsky.social on Thurs March 26 for a lively morning of panels and talks on internet scraping and the future of the open web in the AI age--with perspectives from AI labs, web publishers, scrapers, & a range of experts.

Announcing our latest paper: CommonLID In collaboration with @commoncrawl.bsky.social @mlcommons.org @jhu.edu we built a LID benchmark on actual Common Crawl text covering 109 languages. Existing evaluations overestimate how well LangID works on web data. arxiv.org/abs/2601.18026

CommonLID: Re-evaluating State-of-the-Art Language Identification Performance on Web Data

Language identification (LID) is a fundamental step in curating multilingual corpora. However, LID models still perform poorly for many languages, especially on the noisy and heterogeneous web data of...

arxiv.org

We’re bringing back a Community Spotlight talk series, highlighting cool work being done by members of our community. We’re kicking it off with a talk on running diffusion-based world-models in real time on consumer hardware. Jan 9th at 2 pm US Eastern Time

We are launching a new speaker series at EleutherAI, focused on promoting recent research by our team and community members. Our first talk is by @catherinearnett.bsky.social on tokenizers, their limitations, and how to improve them.

Can you train a performant language model using only openly licensed text? We are thrilled to announce the Common Pile v0.1, an 8TB dataset of openly licensed and public domain text. We train 7B models for 1T and 2T tokens and match the performance similar models like LLaMA 1 & 2

Call for papers! We are organising the 1st Workshop on Multilingual Data Quality Signals with @mlcommons.org and @eleutherai.bsky.social, held in tandem with @colmweb.org. Submit your research on multilingual data quality! Submission deadline is 23 June, more info: wmdqs.org

Today, at 11am ET, @storytracer.org will be giving a live demo on the @mozilla.ai Discord showcasing two Blueprints for creating open datasets: audio transcription using self-hosted Whisper models and document conversion using Docling. Join the event here: discord.com/invite/4jtc8...

Very cool work!

FBK - NLP Research Group@fbk-nlp.bsky.social · last yr.

🚀 **Exciting News!** 🎉 Evalita-LLM is here! 🇮🇹 A new benchmark for evaluating LLMs—offering native Italian tasks, generative challenges, and fair multi-prompt evaluations. Now also available in lm-evaluation harness by @eleutherai.bsky.social ! ArXiv: arxiv.org/abs/2502.02289 #NLProc #LLM #Evaluation

Proud to be at the AI Action Summit representing @eleutherai.bsky.social and the open source community. The focus on AI for the public good is exciting! DM me or @aviya.bsky.social to talk about centering openness, transparency, and public good in the AI ecosystem.

How do a neural network's final parameters depend on its initial ones? In this new paper, we answer this question by analyzing the training Jacobian, the matrix of derivatives of the final parameters with respect to the initial parameters. https://arxiv.org/abs/2412.07003

The latest from our interpretability team: there is an ambiguity in prior work on the linear representation hypothesis: Is a linear representation a linear function (that preserves the origin) or an affine function (that does not)? This distinction matters in practice. arxiv.org/abs/2411.09003

Refusal in LLMs is an Affine Function

We propose affine concept editing (ACE) as an approach for steering language models' behavior by intervening directly in activations. We begin with an affine decomposition of model activation vectors ...

arxiv.org

Proudly building a more open world with @opensource.bsky.social.

Open Source Initiative@opensource.org · 2y ago

The critical question posed to attendees was “How can we best govern and share data to power Open Source AI? https://opensource.org/blog/open-data-and-open-source-ai-charting-a-course-to-get-more-of-both