Excited to share that we will present PICLe at NAACL main conference.

Check out VinaBench, our new #CVPR2025 paper. We introduce a benchmark for faithful and consistent visual narratives. Paper: arxiv.org/abs/2503.20871 Project Page: silin159.github.io/Vina-Bench/

Couldn't attend NAACL HLT 2025 as I didn't get a visa on time 🤷‍♀️ My colleague Mete will present PICLe on my behalf. Feel free to reach out if you want to chat more! paper: arxiv.org/abs/2412.11923 code: github.com/sMamooler/PICLe data: huggingface.co/datasets/smamo…

Thrilled to announce that our work “Fleet of Agents” has been accepted ICML Conference. On average, FoA boosts quality by ~5% while reducing costs to ~40% of SOTA baselines. Blog post after the Neurips deadline ;) Until then: Paper: arxiv.org/abs/2405.06691… Code: github.com/au-clan/FoA

Got ideas for making LLMs more inclusive and culturally aware? 🌍✨ Submit to #MELT Workshop 2025! We’re all about multilingual, multicultural, and equitable LLMs; come share your work! 🧠💬 📅 Oct 10, 2025 | Co-located with Conference on Language Modeling 🔗 melt-workshop.github.io

با بعضی از ایرانی‌هایی که می‌شناسم صحبت کرده‌ام. اما هنوز نمی‌فهمم چرا خیلی‌ها فکر می‌کنن ما نمی‌تونیم چیزی رو عوض کنیم. چی شد که فکر کردیم فقط باید تماشا کنیم و بذاریم دیگران آینده‌مون رو بسازن؟ حقیقت اینه که دنیا رو آدم‌هایی مثل من و تو ساختن. ما می‌تونیم به هر چیزی که ممکنه

To all my fellow countrymen living abroad, Please don’t lose faith. Stay strong. We are a people of peace… descendants of a peaceful nation: Iran. I know the situation is painful and complex. But don’t lose heart. As you wake each day, work harder. Strive for greater success.

🚨New Preprint!! Thrilled to share with you our latest work: “Mixture of Cognitive Reasoners”, a modular transformer architecture inspired by the brain’s functional networks: language, logic, social reasoning, and world knowledge. 1/ 🧵👇

هر کسی که مهارتی دارد، چه داخل ایران چه خارج، و فکر می‌کند می‌تواند کمکی بکند، لطفاً ثبت‌نام کند. با شما در تماس خواهیم بود. docs.google.com/forms/d/1PJAbc…

Both are equally important, and the execution takes much more time and skills. Good-looking ideas and theories come at a dime a dozen (in AI at least), so they're usually worthless until you can confront them with reality through empirical experiments.

NEW PAPER ALERT: Recent studies have shown that LLMs often lack robustness to distribution shifts in their reasoning. Our paper proposes a new method, AbstRaL, to augment LLMs’ reasoning robustness, by promoting their abstract thinking with granular reinforcement learning.

📣The deadline is extended (June 30th)! Consider submitting your multilingual work to #MELTWorkshop2025 #COLM2025

Excited to attend #ACL2025 in Vienna next week! I’ll be presenting our #NAACL2025 paper, PICLe, at the first workshop on Large Language Models and Structure Modeling (XLLM) on Friday. Come by our poster if you’re into NER and ICL with pseudo-annotation. arxiv.org/abs/2412.11923

Also: not the usual #ACL topic, but if you're working on genetic AI models, I’d love to connect! I'm exploring the intersection of behavioral genomics and multi-modal AI for behavior understanding.

🚀 Excited to be at #ACL2025NLP in Vienna! ACL 2025 #ACL2025 Let’s connect and chat about multilingual LLMs & long-context inference! 👋 I’ll be presenting:

Come learn about optimizing in the data space at poster 174!

Heading to Conference on Language Modeling in Montreal? So is WiML! 🎉 We are organizing our first ever event at #CoLM2025 and we want you to choose the format! What excites you the most? Have a different idea? Let us know in the replies! 👇 RT to spread the word! ⏩

The EPFL NLP lab is looking to hire a postdoctoral researcher on the topic of designing, training, and evaluating multilingual #LLMs. (link below) Come join our dynamic group in beautiful Lausanne!

The submission portal is now OPEN to take part in this interesting NeurIPS Conference 2025 data curation competition!! This is the first open-data, open-models, open-source competition for data curation in vision-language reasoning -- learn more 👇 dcvlr-neurips.github.io

The leaderboards lack the needed granularity! When we say model A is better than model B at text, that is too coarse. Even when we say A > B in instruction following, that is still too coarse. I am apparently a tail user as I often find that my ranking of the models for my