ToolWeaver: Weaving Collaborative Semantics for Scalable Tool Use in Large Language Models

2026-01-29

AI Summary

This work addresses the limitations of existing retrieval-based and generative approaches to tool calling: the former struggle to model complex semantics, while the latter assign a unique token to each tool, leading to vocabulary explosion, poor generalization, and difficulty in learning collaborative semantics. To overcome these challenges, the authors propose ToolWeaver, a novel framework that encodes tools as hierarchical sequences, enabling vocabulary size to grow logarithmically with the number of tools. By leveraging dense co-occurrence of shared substructures, ToolWeaver effectively learns collaborative semantics among tools. Integrated with generative alignment fine-tuning, this approach seamlessly incorporates structured tool representations into large language models. Experiments on nearly 47,000 tools demonstrate that ToolWeaver significantly outperforms state-of-the-art methods, achieving breakthroughs in scalability, generalization, and collaborative semantic understanding.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Prevalent retrieval-based tool-use pipelines struggle with a dual semantic challenge: their retrievers often employ encoders that fail to capture complex semantics, while the Large Language Model (LLM) itself lacks intrinsic tool knowledge from its natural language pretraining. Generative methods offer a powerful alternative by unifying selection and execution, tasking the LLM to directly learn and generate tool identifiers. However, the common practice of mapping each tool to a unique new token introduces substantial limitations: it creates a scalability and generalization crisis, as the vocabulary size explodes and each tool is assigned a semantically isolated token. This approach also creates a semantic bottleneck that hinders the learning of collaborative tool relationships, as the model must infer them from sparse co-occurrences of monolithic tool IDs within a vast library. To address these limitations, we propose ToolWeaver, a novel generative tool learning framework that encodes tools into hierarchical sequences. This approach makes vocabulary expansion logarithmic to the number of tools. Crucially, it enables the model to learn collaborative patterns from the dense co-occurrence of shared codes, rather than the sparse co-occurrence of monolithic tool IDs. We generate these structured codes through a novel tokenization process designed to weave together a tool's intrinsic semantics with its extrinsic co-usage patterns. These structured codes are then integrated into the LLM through a generative alignment stage, where the model is fine-tuned to produce the hierarchical code sequences. Evaluation results with nearly 47,000 tools show that ToolWeaver significantly outperforms state-of-the-art methods, establishing a more scalable, generalizable, and semantically-aware foundation for advanced tool-augmented agents.

Problem

• tool use
• semantic representation
• scalability
• large language models
• collaborative semantics

Innovation

• hierarchical tool encoding
• generative tool learning
• structured code tokenization
• collaborative semantics
• scalable tool use