We value your privacy

We use cookies to enhance your browsing experience, serve personalized ads or content, and analyze our traffic. By clicking “Accept”, you consent to our use of cookies.

Blog

Beyond the Model 

Sep 15, 2025

Zero-centered Re-parameterization of LayerNormZero-centered Re-parameterization of LayerNorm

Sep 10, 2025

Language models are getting harder to killLanguage models are getting harder to kill

Aug 18, 2025

What’s actually different about GPT-OSS?What’s actually different about GPT-OSS?

Jul 28, 2025

Can Overparameterization with SVD speed up training?Can Overparameterization with SVD speed up training?

Jul 18, 2025

The Yield Rate of Large Training ClustersThe Yield Rate of Large Training Clusters

Jul 14, 2025

How to calculate evaluation loss correctly (HF Trainer is not correct!)How to calculate evaluation loss correctly (HF Trainer is not correct!)

Jul 8, 2025

Why are Chinese models competitive?Why are Chinese models competitive?

Jun 26, 2025

Beneath the Surface - the Curse of Depth in AI Training ModelsBeneath the Surface - the Curse of Depth in AI Training Models

May 8, 2025

Revisiting LayerNorm: aka Norms are ImportantRevisiting LayerNorm: aka Norms are Important

Apr 15, 2025

Datasets are Where Its AtDatasets are Where Its At

Apr 8, 2025

Three Mistakes Meta Made with Llama4Three Mistakes Meta Made with Llama4

Mar 6, 2025

Introducing Ceramic.aiIntroducing Ceramic.ai

Feb 4, 2025

Cost and Efficiency of DeepSeekCost and Efficiency of DeepSeek

Hey there ai wanderer!

Join our mailing list and get informed on all the latest breakthroughs in AI, machine learning and Ceramic.