Sign in

eleutherai.bsky.social

@eleutherai.bsky.social
2K followers 46 following 45 posts
PostsRepliesMedia
eleutherai.bsky.social @eleutherai.bsky.social · 20/05/2026
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
eleuther.ai
Summer of Open AI Research — EleutherAI
130
eleutherai.bsky.social @eleutherai.bsky.social · 22/04/2026
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!
EleutherAI at ICLR 2026 — where to find our work. Apr 23–27, 2026 in Rio de Janeiro, Brazil. 3 main-track papers (50% acceptance, 3 of 6 submitted) and 1 workshop paper (100%, 1 of 1).
Main track posters, all at Pavilion 4, 10:30 AM – 1:00 PM: (1) "Sparse Autoencoders Trained on the Same Data Learn Different Features" by Gonçalo Paulo and Nora Belrose — Thu Apr 23, board P4-#4004. (2) "Deep Ignorance: Filtering Pretraining Data Builds Tamper-Resistant Safeguards into Open-Weight LLMs" by Kyle O'Brien, Stephen Casper, Quentin Anthony, Tomek Korbak, Robert Kirk, Xander Davies, Ishan Mishra, Geoffrey Irving, Yarin Gal, and Stella Biderman — Fri Apr 24, board P4-#4115. (3) "Evaluating SAE Interpretability without Generating Explanations" by Gonçalo Paulo and Nora Belrose — Sat Apr 25, board P4-#4007.
Workshop paper at the ICBINB Workshop: "Spatial Reasoning is Not a Free Lunch: A Controlled Study on LLaVA" by Nahid Alam, Leema Krishna Murali, Siddhant Bharadwaj, Patrick Liu, Timothy Chung, Drishti Sharma, Akshata A., Kranthi Kiran, Wesley Tam, and Bala Krishna S Vegesna — Mon Apr 27, 13:00–14:25, Room 201C.
030
Reposted by @eleutherai.bsky.social
Kevin Bankston @bankston.bsky.social · 17/03/2026
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...
092
eleutherai.bsky.social @eleutherai.bsky.social · 13/02/2026
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
arxiv.org
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...
12212
eleutherai.bsky.social @eleutherai.bsky.social · 09/01/2026
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
1114
eleutherai.bsky.social @eleutherai.bsky.social · 26/06/2025
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.
1152
eleutherai.bsky.social @eleutherai.bsky.social · 06/06/2025
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
214660
Reposted by @eleutherai.bsky.social
Common Crawl Foundation @commoncrawl.bsky.social · 29/05/2025
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
wmdqs.org
1st Workshop on Multilingual Data Quality Signals
098
eleutherai.bsky.social @eleutherai.bsky.social · 28/04/2025
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...
2105
eleutherai.bsky.social @eleutherai.bsky.social · 24/02/2025
Very cool work!
140
Reposted by @eleutherai.bsky.social
Stella Biderman @stellaathena.bsky.social · 10/02/2025
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.
0234
Reposted by @eleutherai.bsky.social
Nora Belrose @norabelrose.bsky.social · 11/12/2024
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. arxiv.org/abs/2412.07003
49219
eleutherai.bsky.social @eleutherai.bsky.social · 22/11/2024
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
arxiv.org
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 ...
4837
eleutherai.bsky.social @eleutherai.bsky.social · 20/11/2024
Proudly building a more open world with @opensource.bsky.social.
0100