https://magazine.sebastianraschka.com/p/gpt-6-astra-looped-transformers-and

MePo++: Unifying Representation Refinement and Reconciliation for General Continual Learning https://arxiv.org/abs/2609.05075…

Training Large Language Models for Small-Molecule Design with Synthetic Task Scaling https://arxiv.org/abs/2609.04735 https://www.alphaxiv.org/ru/abs/2609.04735

Этим кстати жестко хочется позаниматься https://t.me/researchim/3/7095

https://github.com/Jakub-Nowicki/balatrorm

FrogNano: Training a 4B Coding Agent via Online Task Synthesis https://microsoft.github.io/debug-gym/…