Learning Effective Agentic Organization Policies

Develop algorithms that learn effective organization policies—the strategies for coordinating multiple large language model agents to collaborate concurrently—so that optimal thinking structures are discovered automatically across diverse queries without manual design for each task.

Background

The paper envisions agentic organization, where multiple agents based on LLMs collaborate concurrently to solve complex problems. While parallel thinking methods typically rely on fixed, manually designed workflows, the authors emphasize that adaptivity and dynamism are essential for such systems, and designing optimal structures for every query is intractable.

AsyncThink is introduced as a protocol and training framework (including reinforcement learning) to organize internal thinking via Fork-Join actions. However, the broader challenge of learning effective organization policies in general-purpose settings is explicitly stated as an open problem.

References

Third, learning effective agentic organization policies remains an open problem, as manually designing optimal thinking structures for every possible query is intractable.

— The Era of Agentic Organization: Learning to Organize with Language Models  (2510.26658 - Chi et al., 30 Oct 2025) in Section 1 (Introduction)

Two directions remain open: automatically identifying the required sub-agents and skills from query characteristics rather than curating them manually, and end-to-end sub-agent training with self-evolving adaptation, in which sub-agents are fine-tuned on domain-specific feedback and the framework refines its own planning and routing strategies through experience.

— FinNextAssist: Towards Professional Financial Deep Research Assistant  (2610.03174 - Li et al., 2 Oct 2026) in Conclusion