Dynamic Tool Dependency Retrieval for Lightweight Function Calling

2025-12-18 · Bhrij Patel, Davide Belli, Amir Jalalirad, Maximilian Arnold, Aleksandr Ermolov, Bence Major arxiv

Function calling agents powered by Large Language Models (LLMs) select external tools to automate complex tasks. On-device agents typically use a retrieval module to select relevant tools, improving performance and reducing context length. However, existing retrieval methods rely on static and limited inputs, failing to capture multi-step tool dependencies and evolving task context. This limitation often introduces irrelevant tools that mislead the agent, degrading efficiency and accuracy. We propose Dynamic Tool Dependency Retrieval (DTDR), a lightweight retrieval method that conditions on both the initial query and the evolving tool calling plan. DTDR models tool dependencies from function calling demonstrations, enabling adaptive retrieval as plans unfold. We benchmark DTDR against state-of-the-art retrieval methods across multiple datasets and LLM backbones, evaluating retrieval precision, downstream task accuracy, and computational efficiency. Additionally, we explore strategies to integrate retrieved tools into prompts. Our results show that DTDR improves function calling success rates between $23\%$ and $104\%$ compared to state-of-the-art static retrievers.

📄 PDF Abstract BibTeX arXiv:2512.17052

Code (0)

등록된 구현이 없습니다.

Tasks

Computational Efficiency

ProTIP: Progressive Tool Retrieval Improves Planning

2023-12-16 · Raviteja Anantha, Bortik Bandyopadhyay, Anirudh Kashi, Sayantan Mahinder 외

Large language models (LLMs) are increasingly employed for complex multi-step planning tasks, where the tool retrieval (TR) step is crucial for achieving successful outcomes. Two prevalent approaches for TR are single-st…

Contrastive Learning Retrieval

Bidirectional Semantic Complementary Tool Retrieval for Remote Sensing Agents

2026-04-29 · Zeyuan Wang, Dongyang Hou, Cheng Yang, Xuezhi Cui 외 arxiv

Large language model (LLM)-based agents provide a novel paradigm for the automated processing of remote sensing(RS) data. Their success in complex RS tasks rely on extensive specialized tool libraries. However, tool docu…

Natural Language Queries Continual Learning

Towards Reliable Benchmarking: A Contamination Free, Controllable Evaluation Framework for Multi-step LLM Function Calling

2025-09-30 · Seiji Maekawa, Jackson Hassell, Pouya Pezeshkpour, Tom Mitchell 외 arxiv

Existing benchmarks for tool-augmented language models (TaLMs) lack fine-grained control over task difficulty and remain vulnerable to data contamination. We present FuncBenchGen, a unified, contamination-free framework …

Dynamic Context Tuning for Retrieval-Augmented Generation: Enhancing Multi-Turn Planning and Tool Adaptation

2025-06-05 · Jubin Abhishek Soni, Amit Anand, Rajesh Kumar Pandey, Aniket Abhishek Soni

Retrieval-Augmented Generation (RAG) has significantly advanced large language models (LLMs) by grounding their outputs in external tools and knowledge sources. However, existing RAG systems are typically constrained to …

RAG Retrieval Retrieval-augmented Generation

Beyond Single-Shot: Multi-step Tool Retrieval via Query Planning

2026-01-12 · Wei Fang, James Glass arxiv

LLM agents operating over massive, dynamic tool libraries rely on effective retrieval, yet standard single-shot dense retrievers struggle with complex requests. These failures primarily stem from the disconnect between a…

Zero-shot Generalization Reinforcement Learning