Periodic Labs' Neon model surpasses GPT-6 Astra on materials science benchmark
Exciting results, @LiamFedus! Congrats to the whole team at Periodic Labs!
For those interested in advanced materials science applications, Neon's achievement shows a new open-source model can compete with proprietary models on specialized benchmarks.
Periodic Labs announced that their open-source model Neon, trained using 1,300 H200s and experimental data, outperformed GPT-6 Astra on an analysis benchmark for materials science. The team built high-throughput labs in Menlo Park to create a loop between experiments and models. Neon is focused on solving hard problems like superconductors and magnets.
Exciting results, @LiamFedus! Congrats to the whole team at Periodic Labs!
Exciting results, @LiamFedus ! Congrats to the whole team at Periodic Labs! Liam Fedus @LiamFedus We built high-throughput materials labs in Menlo Park to create a loop between experiments and models. The labs generate fresh data, the models learn from it, and then help us decide what to try next. Using only 1,300 H200s, plus months of our experimental data, we mid-trained and RL’d an open-source model to surpass GPT-6 Astra on our analysis benchmark. We call it Neon. This is real footage from our lab. We’re focusing first on hard problems in materials science, including superconductors, magnets, and semiconductor materials. Read our blog posts below. Your browser does not support the video tag. 🔗 View on Twitter 🔗 View Quoted Tweet 💬 6 🔄 9 ❤️ 127 👀 18293 📊 17 ⚡
