AI RESEARCH
Milestone-Guided Policy Learning for Long-Horizon Language Agents
arXiv CS.CL
•
ArXi:2605.06078v1 Announce Type: new While long-horizon agentic tasks require language agents to perform dozens of sequential decisions