Can LLM Agents Sustain Long-Horizon Organizational Dynamics?
This work addresses the challenge of maintaining coherent, long-horizon behavior in LLM-based social simulations for organizational dynamics, which is a key step toward building reliable organizational simulators.
The paper investigates whether LLM agents can sustain coherent behavior in long-horizon organizational simulations, proposing TaskWeave, a hierarchical framework that uses a Formulate-Partition-Diagnose-Align cycle and dependency-aware trace memory. In a year-long IT company simulation, TaskWeave outperformed other multi-agent frameworks in organizational coherence, execution grounding, and downstream utility.
Large language agents are increasingly used for social simulation, yet it remains unclear whether they can sustain coherent behavior in structured organizations, where goals must propagate through hierarchy, tasks depend on prior execution, and artifacts accumulate over long horizons. We formulate long-horizon organizational simulation as a memory-centered coordination problem and introduce TaskWeave, a hierarchical agentic framework that maintains planning states through a Formulate-Partition-Diagnose-Align cycle and grounds execution through dependency-aware trace memory. We evaluate TaskWeave in a year-long IT company simulation and compare it with other multi-agent frameworks on organizational coherence, execution grounding, and downstream enterprise NLP utility. Experiments show that TaskWeave supports coherent and long-horizon organizational dynamics while producing grounded artifacts and adapting to external environments. These findings suggest that structured simulation memory is a key mechanism for building reliable LLM-based organizational simulators.