Congrats again to our brilliant students @davidguzman1120 @_yongjinny for receiving the "Oral Paper Award" at the #ACL2025NLP Workshop on Research on Agent Language Models (REALM)! Check out how Reasoning LLMs optimize self interests over collective success 📊 in our paper

📚Paper : https://t.co/DZp9CgJoRe 💻Project :

🎉 Excited to share that our paper, "Automated Filtering of Human Feedback Data for Aligning Text-to-Image Diffusion Models", will be presented at #ICLR2025!​ 📅 Date: April 24 🕒 Time: 10 am 📍 Location: Hall 3 + Hall 2B #152 Come check out our poster!

With o1-like models charging by token usage 💸, we want to avoid verbal redundancy. Our self-training method can elicit concise reasoning, reducing token usage by 30% 🔥 while maintaining accuracy! w/ MS students @Tergel_Munkhbat and @shkimsally at https://t.co/CeSggg6BR9 🚀

Thrilled to announce our #EMNLP2023 Findings paper on explainable hate speech detection using LLM-generated data! 🎉📄 A huge thanks to @itsnamgyu for sharing our work😊

I asked my audience: “If you could only listen to one podcast for the next 5 years, what would it be?” Here are 10 of the most popular replies:

Everybody writes (texts, work emails, sales copy, etc.) But few people write well. Here are 7 free, must-have writing tools:

If there is one thing the deep learning revolution has taught us, it's that neural nets will outperform hand-designed heuristics, given enough compute and data. But we still use hand-designed heuristics to train our models. Let's replace our optimizers with trained neural nets!

All the educational resources to learn Machine Learning in one spot:

Colleges keep publishing their machine learning courses online. 100% free: ↓ - MIT 6.S191 Introduction to Deep Learning - DS-GA 1008 Deep Learning - UC Berkeley Full Stack Deep Learning - UC Berkeley CS 182 Deep Learning - Cornell Tech CS 5787 Applied Machine Learning

I've been trading for 10 years. Everything you need to know is in these 25 threads:

Tired of tuning your neural network optimizer? Wish there was an optimizer that just worked? We’re excited to release VeLO 🚲, the first hyperparameter-free learned optimizer that outperforms hand-designed optimizers on real-world problems: https://t.co/zarCWuqIWb 🧵

We’ve released ImageNetX: a set of human annotations for the popular ImageNet benchmark to gauge model robustness strengths/weaknesses — one of the first large scale efforts to pinpoint mistake types in AI computer vision systems. Explore the dataset ⬇️