We post-train LLMs for math, for code, for instruction-following. Why not for scientific discovery?
🫎 MOOSE-Star (ICML 2026) : the first scalable SFT recipe for discipline-agnostic scientific hypothesis discovery.
By @Yang_zy223 & @LidongBing from MiroMind.
Hello from #ICML2025#! 👋
Together with @GoogleResearch, we’re presenting over 140 papers, as well as hosting workshops, talks and demo sessions.
Check out our schedule. →
Come say hi at #ICML2024#! 👋
We’ll be hosting live research demos on Gemini Nano, LearnLM - our family of models for education - and TacticAI, an AI assistant for football.
You can also check out our oral, spotlight and poster presentations. →