Tell me you use cursor without telling me you use cursor. #cursor

Happy to announce "Performative Prediction on Games and Mechanism Design" was accepted at AISTATS Conference 2025, and got spotlight at HAIC(ICLR 2026 workshop) with Mehrnaz Mofakhami Fernando P. Santos Gauthier Gidel Simon Lacoste-Julien (Mila and UvA) arxiv.org/abs/2408.05146 Details below 1/9🧵

Scientific discovery with LLMs has so much potential yet is underexplored. Our new benchmark **LLM-SRBench** enable rigorous evaluations of equation discovery with LLMs! 🧠Key takeaway: Even SOTA discovery models with strong LLM backbones still fail to discover mathematical

Happy to host Karan tomorrow to talk about their work on one-minute video generation with TTT.

We'll be presenting the Forgetting Transformer during the poster session at 3 pm on April 25th at #ICLR2025 (board number 282). Come and chat with us! • Poster info: iclr.cc/virtual/2025/p… • Paper: arxiv.org/abs/2503.02130 • Code: github.com/zhixuan-lin/fo…

Sharing our poster ICLR 2026! Please stop by poster if you're interested 😊 Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA 📍 Hall 3 + Hall 2B, Poster #262 🗓️ Saturday, April 26, 10:00 a.m. — 12:30 p.m.

I'm presenting our recent work on "Pitfalls of Memorization" today at ICLR Number #304 at 3pm.. Come say hi! iclr.cc/virtual/2025/p…

Excited to share our latest work on EvoTune, a novel method integrating LLM-guided evolutionary search and reinforcement learning to accelerate the discovery of algorithms! 1/12🧵

🚀 Check out our recent work on algorithm discovery with LLMs, leveraging evolutionary search and RL fine-tuning! It's been an incredible and enjoyable experience working with such a talented team and amazing advisors🌟! Read Anja’s 🧵and head to claire-labo.github.io/EvoTune for more!

Parallel computing is beautiful. It still amazes me — you run a command, go to sleep, and the machines do the work for you.

Thrilled to share that our paper, "Autoencoder-Based Hybrid Replay for Class-Incremental Learning," has been accepted at ICML 2025! In this work, we propose a novel Hybrid Autoencoder (HAE) architecture designed to tackle catastrophic forgetting in class-incremental learning. By

Deliberate practice is accepted to #ICML2025 as a spotlight (top 2.6%!) 🚀

"Compositional Risk Minimization" Tackling unseen attribute combinations with additive energy models at #ICML2025. 💡 Check out the excellent summary by Divyat Samadhiya

Idea: Merging generative verification and solution generarion during RL training of LLM reasoners. Why? This allows you to scale inference compute both sequentially (long CoT) and parallel (Best of N, weighted maj voting). Next? Generation and verification to be trained end

Today, we celebrate International Women in Mathematics Day in memory of Maryam Mirzakhani, the first woman to win the Fields Medal (2014), to honor the contributions and achievements in women in mathematics in hopes to inspire future generations.

Happy to announce that we (w/ my student Daking Rai ) will present a tutorial on 𝐌𝐞𝐜𝐡𝐚𝐧𝐢𝐬𝐭𝐢𝐜 𝐈𝐧𝐭𝐞𝐫𝐩𝐫𝐞𝐭𝐚𝐛𝐢𝐥𝐢𝐭𝐲 𝐟𝐨𝐫 𝐋𝐚𝐧𝐠𝐮𝐚𝐠𝐞 𝐌𝐨𝐝𝐞𝐥𝐬! Look forward to meeting people ICML Conference Stay tuned! ziyu-yao-nlp-lab.github.io/ICML25-MI-Tuto… George Mason University Department of Computer Science - George Mason

Nature:

🚨 Preprint Alert 🚀 📄 seq-JEPA: Autoregressive Predictive Learning of Invariant-Equivariant World Models arxiv.org/abs/2505.03176 Can we simultaneously learn both transformation-invariant and transformation-equivariant representations with self-supervised learning (SSL)?