Financial Assistance applications are now open! If you face financial barriers to attending ICLR 2025, we encourage you to apply. The program offers prepay and reimbursement options. Applications are due March 2nd with decisions announced March 9th. iclr.cc/Conferences/20…

Last year, we funded 250 authors and other contributors to attend #ICLR2024 in Vienna as part of this program. If you or your organization want to directly support contributors this year, please get in touch! Hope to see you in Singapore at #ICLR2025!

AABI is back! This year, for the first time in Asia (co-located with ICLR in Singapore). Send us all your new probabilistic ML papers, the deadline is just a few days after the ICML one!

if you would like to attend #ICLR2025 but have financial barriers, apply for financial assistance! our priority categories are student authors, and contributors from underrepresented demographic groups & geographic regions. deadline is march 2nd.

Check out MODEL SELECTOR, a framework for label-efficient selection of pretrained classifiers. We reduce the labeling cost by up to 94.15% to identify the best model. Joint work with Andreas Kirsch 🇺🇦 Torsten Hoefler 🇨🇭 Andreas Krause Nezihe Merve Gürel (nmervegurel.bsky.social) accepted at #AISTATS2025 !

Super excited about our ICLR workshop on synthetic data: "Will Synthetic Data Finally Solve the Data Access Problem?" Submit your work by Feb 5 AoE and join us in Singapore 🎉: synthetic-data-iclr.github.io #ICLR2025 #SyntheticData

Looking forward to our sparse LLMs workshop ICLR 2026 for a deep dive into sparsity, mixture of experts, quantization, hardware, and inference for LLMs. Submission deadline is February 7th!

workshops ICLR 2026 selection process bts blog.iclr.cc/2025/02/03/wor… p.s. workshops soon close for submissions, send them your exciting works!

My first PhD work, which accelerates LLM inference TTFT! Thanks a lot for my awesome advisor Ce Zhang and colab advisor Beidi Chen! Speculative Prefill is a training-free framework that achieves up to 7x maximal QPS with Llama 3.1 405B on vLLM! arxiv.org/abs/2502.02789

We are seeking Volunteer & Mentorship Chairs for #ICLR2025. If interested, fill out: forms.gle/SwhAMQJCx5iwNe… Initial work (signing up volunteers & senior mentors) will start soon. Bulk of work is closer to conference (assigning volunteers to tasks, office hours for mentors).

Can Reasoning Improve Safety & Security? Red-Teaming Analysis for Claude 3.7 🚀 Claude 3.7 Sonnet Thinking: A New Era of Hybrid Reasoning? Anthropic's latest release introduces a Thinking mode, letting users switch between rapid responses and step-by-step reasoning. But does

A dark time for democracy in Türkiye. The order to detain Istanbul Mayor Ekrem İmamoğlu is the latest example of the justice system being weaponized to cancel the democratic process. His detention violates the rights of millions of Istanbul voters.

Announcing the keynote speakers for #ICLR2025! Speakers will cover topics ranging from foundational advances in language models, AI safety, open-ended learning, and the nature of intelligence itself. blog.iclr.cc/2025/04/11/ann…

I want to share a blog post about our paper "All Models are Wrong, Some are Useful: Model Selection with Limited Labels", which we will present at AISTATS Conference 2025 next week With Patrik Okanovic, Jannes Kasper, Torsten Hoefler 🇨🇭, Andreas Krause, Nezihe Merve Gürel (nmervegurel.bsky.social)

Caglar Gulcehre and I are excited to share our lecture slides for EE-628 Training Large Language Models course: epfl.ch/labs/lions/tea… If you have any feedback, please reach out to us. I am also at #ICLR25.

Huge thanks to the #ICLR2025 Organizing Committee (including many who couldn't make it to the conference) 👏👏👏

Speculative Prefill got accepted by #ICML2025! Thanks Ce Zhang Beidi Chen!! Significantly improves LLM prefill throughput and latency with up to 405B models without training! We achieved 2-7x higher QPS on vLLM servers while preserving quality. github.com/Jingyu6/specul…

Happening now at AISTATS Conference, our paper at the poster session now with Patrik Okanovic presenting: "All models are wrong, some are useful: Model Selection with Limited Labels"

Patrik Okanovic and Andreas Kirsch 🇺🇦 have presented the paper "All models are wrong, some are useful: Model Selection with Limited Labels" at AISTATS Conference! 👉arxiv.org/pdf/2410.13609 👉github.com/RobustML-Lab/m… Work of Patrik Okanovic, Andreas Kirsch 🇺🇦, Jannes Kasper, Torsten Hoefler 🇨🇭, Andreas Krause, Nezihe Merve Gürel (nmervegurel.bsky.social)

If you are working on AI agents-related research and want to share with the community, please consider submitting to our AI Agent workshop at COLM!!!🤖🎉 Submission Deadline: June 23rd, 2025 Workshop date: October 10th, 2025