一作ACL 2026Poster· Natural Language Processing
KARL : Reinforcement Learning for LLM Agents on Multi-Turn Knowledge-Intensive Agentic Tasks
Xueqiao Sun · Xiao Liu · Bowen Lv · Hanchen Zhang · Bohao Jing · Zehan Qi · Yifan Xu · Yuxiao Dong · Jie Tang
摘要
Large Language Models have shown remarkable potential as autonomous agents, but their effectiveness in knowledge-intensive tasks remains limited by passive knowledge utilization. We introduce KARL (Knowledge-Augmented Reinforcement Learning), a framework that enables LLM agents to dynamically exp...