Fetching the paper…

LongAgent: Scaling Language Models to 128k Context through Multi-Agent Collaboration · Around