Tulip Agent -- Enabling LLM-Based Agents to Solve Tasks Using Large Tool Libraries
This addresses the bottleneck of high inference costs and limited scalability for autonomous agents using large tool sets, though it is an incremental improvement over existing methods.
The paper tackles the problem of enabling LLM-based agents to efficiently use large tool libraries by introducing Tulip Agent, an architecture that avoids encoding all tool descriptions in the prompt, reducing inference costs and allowing scalability, as demonstrated in mathematics and robotics contexts.
We introduce tulip agent, an architecture for autonomous LLM-based agents with Create, Read, Update, and Delete access to a tool library containing a potentially large number of tools. In contrast to state-of-the-art implementations, tulip agent does not encode the descriptions of all available tools in the system prompt, which counts against the model's context window, or embed the entire prompt for retrieving suitable tools. Instead, the tulip agent can recursively search for suitable tools in its extensible tool library, implemented exemplarily as a vector store. The tulip agent architecture significantly reduces inference costs, allows using even large tool libraries, and enables the agent to adapt and extend its set of tools. We evaluate the architecture with several ablation studies in a mathematics context and demonstrate its generalizability with an application to robotics. A reference implementation and the benchmark are available at github.com/HRI-EU/tulip_agent.