1/9 Current optimizers are static during training. What if the optimizer could think and react while the network trains? We let an agent watch the training dynamics and rewrite the optimizer's source code as training goes: a reactive optimizer. In every language and image setting we tried that has no tuned recipe, it matches or beats the best recipe one can one-shot, and it guards against loss spikes and overfitting by itself. In the near future, optimization decisions may belong to agents, not humans. 📝 https://xingjianbai.com/reactive-optimizers/ 💻 Try it on your task: https://github.com/xingjian-bai/opal 🧵

Gemini 4 Argon is available now for Google AI Ultra and Pro subscribers in @GeminiApp and AI Mode in Search, Ultra subscribers in @Antigravity, and to developers and enterprises via the Gemini API in @GoogleAIStudio and the Gemini platform. Argon will launch at an introductory price of $2 per million input tokens and $10 per million output tokens, with cached input tokens priced at 95% off input token price.