AI RESEARCH
Enhancing Multi-Agent Communication through Attention Steering with Context Relevance
arXiv CS.AI
•
ArXi:2605.30136v1 Announce Type: new LLM-based multi-agent systems have nstrated remarkable performance on complex tasks through collaborative reasoning. However, these systems tend to rapidly accumulate extremely long conversation histories during interaction. As conversations lengthen, relevant information is increasingly diluted by irrelevant context, leading to degraded performance. In this work, we present Agent-Radar, a