Periodic Labs pushes Kimi 2.5 base model past Astra with specialized scientific training
teortaxesTex · x · 2026-09-16
Liam Fedus (Periodic Labs) describes how Neon was built from three pieces—labs, research, and infra—using unique experimental data and efficient infrastructure to train specialized scientific models on modest compute. Commenters note Periodic Labs pushed the Kimi 2.5 base model to exceed even Astra.
Related event: Periodic Labs beats GPT-6 on materials science with just 1,300 H200s(41 posts)→
More from Models
- Cartesia's new voice model family draws praise; WER alone can't capture context-correct speech — buckymoore · 2026-09-16
- Replication of no-CoT evals shows GPT-Astra makes a qualitative jump across all datasets — dhadfieldmenell · 2026-09-16
- Burkov maxed out Codex's $200/month plan and switched back to Claude Astra — ruslansv · 2026-09-16
- DeepSeekMath-V2 makes verification the product, scaling verifier compute ahead of the generator — le_james94 · 2026-09-16
- ChatGPT Plus Work Projects Bug Persists for Days While OpenAI Marks It Resolved — OnwardUpwardForward · 2026-09-16
- Google reportedly building math-specialized Gemini DeepThink Mathematica model — basedjensen · 2026-09-16