I'm excited to announce that my nonfiction book, "Lost in Automatic Translation: Navigating Life in English in the Age of Language Technologies", will be published this summer by Cambridge University Press. I can't wait to share it with you! 📖🤖 cambridge.org/core/books/los…

🚀 Tower+: our latest model in the Tower family — sets a new standard for open-weight multilingual models! We show how to go beyond sentence-level translation, striking a balance between translation quality and general multilingual capabilities. 1/5 arxiv.org/pdf/2506.17080

Super grateful to share that our work has been accepted as #ACL2025 oral presentation 🍀✨ See you in Vienna! 🇦🇹

Why should you attend this talk? 🤔 A. Nishant put so much effort B. Learn the real limitations of MCQA C. Great takeaways for building better benchmarks D. All of the above ✔️

📢When LLMs solve tasks with a mid-to-low resource input/target language, their output quality is poor. We know that. But can we pin down what breaks inside the LLM? We introduce the 💥translation barrier hypothesis💥 for failed multilingual generation. arxiv.org/abs/2506.22724

CLIPPER has been accepted to #COLM2025! In this work, we introduce a compression-based pipeline to generate synthetic data for long-context narrative reasoning tasks. Excited to be in Montreal this October🍁

You have a budget to human-evaluate 100 inputs to your models, but your dataset is 10,000 inputs. Do not just pick 100 randomly!🙅 We can do better. "How to Select Datapoints for Efficient Human Evaluation of NLG Models?" shows how.🕵️ (random is still a devilishly good baseline)

📣Thrilled to announce the drop of EXAONE 4.0, the next-generation hybrid AI. 🙌Prepare to be amazed by EXAONE’s capabilities. #EXAONE #LG_AI_Resrarch #HybridAI #AI lgresearch.ai/blog/view?seq=…

(Repost due to mistaken deletion😢): Evaluating topic models (& doc clustering methods) is hard. In fact, since our paper critiquing standard eval practices 4 years ago, there hasn't been a good replacement metric That ends today! Our ACL paper introduces a new evaluation🧵

Maybe don't use an LLM for _everything_? Last summer, I got to fiddle again with content diversity Adobe Research Adobe and we showed that agentic pipelines that mix LLM-prompt steps with principled techniques can yield better, more personalized summaries

I will be presenting our work 𝗠𝗗𝗖𝘂𝗿𝗲 at #ACL2025NLP in Vienna this week! 🇦🇹 Come by if you’re interested in multi-doc reasoning and/or scalable creation of high-quality post-training data! 📍 Poster Session 4 @ Hall 4/5 🗓️ Wed, July 30 | 11-12:30 🔗 aclanthology.org/2025.acl-long.…

I'm at #ACL2025 presenting our work on enhancing equitable cultural alignment through multi-agent debate ✨ Come visit our oral presentation! 📍Computational Social Science and Cultural Analytics session (Level 1 1.85) 📆Tuesday (7/29) 2-3:30pm 📝aclanthology.org/2025.acl-long.…

I'll also be presenting our paper on using question-answer pairs as a new signal for spotting translation errors 🕵️ Come to talk more about MT evaluation! 📍Poster session (Hall X4, X5) 📆Tuesday (7/29) 4-5:30pm 📝aclanthology.org/2025.findings-…

Today, we're releasing The Circuit Analysis Research Landscape: an interpretability post extending & open sourcing Anthropic's circuit tracing work, co-authored by Paul Jankura, Google DeepMind, Goodfire EleutherAI, and Decode Research. Here's a quick demo, details follow: ⤵️

🚀New paper alert! 🚀 In our work AI at Meta we dive into the struggles of mixing languages in largely multilingual Transformer encoders and use the analysis as a tool to better design multilingual models to obtain optimal performance. 📄: arxiv.org/abs/2508.02256 🧵(1/n)

AI shows ingroup bias towards AI content! If we deploy LLMs in decision-making roles (e.g., purchasing goods, selecting academic submissions) they will favor LLM agents over ordinary humans pnas.org/doi/10.1073/pn…

🚀New dataset release: WildChat-4.8M 4.8M real user-ChatGPT conversations collected from our public chatbots: - 122K from reasoning models (o1-preview, o1-mini): represent real uses in the wild and very costly to collect - 2.5M from GPT-4o 🔗 hf.co/datasets/allen… (1/4)

EPFL , ETH Zurich and #CSCS today released Apertus, Switzerland's first large-scale, multilingual language model (LLM). As a fully open LLM, it serves as a building block for developers and organizations to create their own applications: cscs.ch/science/comput… #Apertus #AI