🚨 2 days until the #CoLLAs2024 abstract deadline! 🚨 Stay ahead and complete the abstract submission today! 👇 openreview.net/group?id=lifel…

📢📢📢The abstract deadline for #CoLLAs2024 is tomorrow 👇

🚨 Deadline Alert for #CoLLAs2024: 2 Days Left! 🚨 🔥 Don’t miss out on the opportunity to showcase your work to a global audience of lifelong machine learning experts! 🚀 Submit your work to the Third International Conference on Lifelong Learning Agents by February 15th AOE 👇

Roses are Red, Violets are Blue; Don’t Forget to Submit, (Tomorrow) on OpenReview!

Don't forget - the deadline for #CoLLAs2024 is tomorrow! 💘👇

Violets are blue Roses are red Missed ICML? Try CoLLAs instead :)

Did you know that you can present your recently published top-tier journal papers at #CoLLAs2024? 🤯 Learn more 👇

🚨Deadline extension for CLVISION 2024🚨 📢 Submit before March 5th sites.google.com/view/clvision2…

🚨 Interested in the latest and greatest in Continual Learning research!? 🚀 All videos from our full 24-hour ContinualAI Unconference (October 2023) are now available on YouTube! Check them out today! 👇 youtube.com/playlist?list=… Website: unconf.continualai.org

I was fortunate to receive a Georgia Tech Computing outstanding graduate research award during the last year of my PhD - thanks to all of my wonderful collaborators and advisor, Zsolt Kira! 🎉

Join us at the #CVPR2025 Workshop on #ContinualLearning in #ComputerVision the 18th of June in Seattle🌳! Plenty of chances of networking and an exciting program for you!🎊 🔗sites.google.com/view/clvision2… 🧵(1/4)

Low-rank decomposition for LLM compression adds significant overhead by splitting 1 matrix into 2. My gang's *MoDeGPT* fixes this by decomposing 2 adjoint matrices! By leveraging decomposition-compression duality at the modular level, we achieve SOTA training-free compression!

I’m hiring 2 fully-funded PhD students to join our new Lifelong ML lab at the University of Bremen, where our mission is to make ML/AI more sustainable, adaptive, robust & inclusive. Application deadline Dec. 10: uni-bremen.de/en/university/… Please share & reach out with questions!

Visit #NeurIPS papers from RIPL! Includes robust finetuning of FMs, diffusion models for action, and VLA action tokenization. I'll also be giving a (remote) talk on continual learning. buff.ly/49BY9K9 buff.ly/3OQd5ef buff.ly/41wc1DL @icatgt Machine Learning at Georgia Tech

Today is my last day at Samsung Research America. I’m so grateful for the talented colleagues, exciting projects, and incredible mentorship from Yen-Chang Hsu that shaped the start of my career. Thank you, SRA, for this rewarding chapter—excited for what’s next!

🔥🔥Thrilled to share our work ICLR 2026🔥🔥 🤔 Low-rank decomposition for LLM compression adds significant overhead by splitting 1 matrix into 2. 🔑 Key contribution: ✅ Introduction of decomposition-compression duality: By leveraging the duality, we compress two adjoint

Excited to share some research updates! ➡️ MoDeGPT @ ICLR 2025: Modular Decomposition for LLM Compression (led by Chi-Heng Lin 貓貓🐱). Post below 👇 ➡️ FlexiGPT @ NAACL 2025: Pruning LLMs with low-rank weight sharing. Learn more: arxiv.org/abs/2501.14713

🚀 ICLR 2025 Oral Presentation! 🚀 Exciting to see our work on modular LLM compression has been selected for an Oral at #ICLR2025! 🎉🎉 If you missed it, check out our key contribution: ⬇️⬇️ #LLM #ModelCompression #MatrixDecomposition