Search papers, labs, and topics across Lattice.
The paper introduces Hybrid Toolset Agentization & Adaptation (HTAA), a hierarchical framework designed to improve the efficiency and reliability of large language models using numerous tools. HTAA reduces the planner's action space by encapsulating frequently co-used tools into specialized agent tools and uses Asymmetric Planner Adaptation for effective coordination between the high-level planner and agent tools. Experiments on a real-world dataset and benchmarks demonstrate that HTAA achieves higher task success rates, shorter tool calling trajectories, and reduced context overhead, with successful production deployment.
LLMs can navigate complex tool-use scenarios more effectively by organizing tools into specialized agents and aligning high-level planning with agent execution.
Enabling large language models to scale and reliably use hundreds of tools is critical for real-world applications, yet challenging due to the inefficiency and error accumulation inherent in flat tool-calling architectures. To address this, we propose Hybrid Toolset Agentization & Adaptation (HTAA), a hierarchical framework for scalable tool-use planning. We propose a novel toolset agentization paradigm, which encapsulates frequently co-used tools into specialized agent tools, thereby reducing the planner's action space and mitigating redundancy. To ensure effective coordination, we design Asymmetric Planner Adaptation, a trajectory-based training paradigm that aligns the high-level planner with agent tools via backward reconstruction and forward refinement. To validate the performance of HTAA, we conduct experiments on a real-world internal dataset, InfoVerify, based on the POI validation workflow of China's largest online large-scale ride-hailing platform, featuring long-horizon executable tool trajectories. Experiments on InfoVerify and widely-used benchmarks show that HTAA consistently achieves higher task success rates, requires short tool calling trajectories, and significantly reduces context overhead compared to strong baselines. Furthermore, in a production deployment, HTAA substantially reduces manual validation effort and operational cost, demonstrating its practical efficacy.
