The Cognitive Bandwidth Bottleneck: Shifting Long-Horizon Agent from Planning with Actions to Planning with Schemas

2025-10-08

AI Summary

In open-world, long-horizon tasks, combinatorial explosion of the action space renders action-level planning (PwA) computationally intractable and non-scalable. Method: This paper proposes shifting from PwA to schema-based planning (PwS), grounded in cognitive bandwidth theory, to quantitatively characterize the critical transition point in action representation selection and reveal the coupling between model capacity and action representation. We conduct controlled experiments comparing PwA and PwS across varying action-space scales and model capabilities. Results: PwS significantly improves long-horizon planning efficiency and scalability—especially under large action spaces—and yields concrete design principles for efficient PwS agents. This work provides both theoretical foundations and practical guidelines for building scalable autonomous agents.

Technology Category

• Multiagent Systems: Multiagent Planning
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience

Abstract

Enabling LLMs to effectively operate long-horizon task which requires long-term planning and multiple interactions is essential for open-world autonomy. Conventional methods adopt planning with actions where a executable action list would be provided as reference. However, this action representation choice would be impractical when the environment action space is combinatorial exploded (e.g., open-ended real world). This naturally leads to a question: As environmental action space scales, what is the optimal action representation for long-horizon agents? In this paper, we systematically study the effectiveness of two different action representations. The first one is conventional planning with actions (PwA) which is predominantly adopted for its effectiveness on existing benchmarks. The other one is planning with schemas (PwS) which instantiate an action schema into action lists (e.g., "move [OBJ] to [OBJ]" -> "move apple to desk") to ensure concise action space and reliable scalability. This alternative is motivated by its alignment with human cognition and its compliance with environment-imposed action format restriction. We propose cognitive bandwidth perspective as a conceptual framework to qualitatively understand the differences between these two action representations and empirically observe a representation-choice inflection point between ALFWorld (~35 actions) and SciWorld (~500 actions), which serve as evidence of the need for scalable representations. We further conduct controlled experiments to study how the location of this inflection point interacts with different model capacities: stronger planning proficiency shifts the inflection rightward, whereas better schema instantiation shifts it leftward. Finally, noting the suboptimal performance of PwS agents, we provide an actionable guide for building more capable PwS agents for better scalable autonomy.

Problem

• Optimizing action representation for long-horizon agents in combinatorial environments
• Comparing planning with actions versus schemas for scalable autonomy
• Identifying inflection points where schema-based planning outperforms action-based methods

Innovation

• Shifts planning from actions to schemas
• Uses cognitive bandwidth framework for analysis
• Identifies inflection point for representation choice