Agentic Multi-Turn Reasoning: A Fairness Approach.

arXiv:2609.33323v1 Announce Type: new Abstract: Recent advances in Large Language Models (LLMs) have enabled agentic systems capable of solving complex tasks through multi-turn planning, tool use, verification, and memory updates. However, learning agentic sys