Migel Tissera(@migtissera):This is awesome. Periodic fine-tuned Kimi-K2.6, and the model beats Fable-5.1 and GPT-6-Astra in their own scientific work. This is why open source is important and also one of the reasons that Anthropic is so loud for regulatory capture. They will have no moat.

2026.09.15 17:01

This is awesome. Periodic fine-tuned Kimi-K2.6, and the model beats Fable-5.1 and GPT-6-Astra in their own scientific work. This is why open source is important and also one of the reasons that Anthropic is so loud for regulatory capture. They will have no moat.

Liam Fedus@LiamFedus

2026.09.15 16:20

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.