а пока хз использую из торча их мюон, нифига

Building a web search engine from scratch in two months with 3 billion neural embeddings https://blog.wilsonl.in/search-engine/

https://www.inceptionlabs.ai/blog/introducing-mercury-2

On Data Engineering for Scaling LLM Terminal Capabilities https://arxiv.org/abs/2602.21193 https://www.alphaxiv.org/ru/overview/2602.21193…

из-за матетематики или в коде трабл?

CodeScaler: Scaling Code LLM Training and Test-Time Inference via Execution-Free Reward Models https://arxiv.org/abs/2602.17684…