Search papers, labs, and topics across Lattice.
The paper introduces Tool-DC, a divide-and-conquer framework with training-free (TF) and training-based (TB) variants, to improve LLM tool-calling performance in long-context scenarios with many candidate tools. Tool-DC leverages a "Try-Check-Retry" paradigm to reduce reasoning complexity and exploit LLM self-reflection. Experiments on BFCL and ACEBench show Tool-DC (TF) achieves up to +25.10% improvement over baselines, and Tool-DC (TB) allows smaller models like Qwen2.5-7B to match or exceed the performance of larger proprietary models.
LLMs can now navigate massive toolsets with a "Try-Check-Retry" loop, boosting tool-calling accuracy by up to 25% and letting smaller models punch above their weight.
Tool-calling empowers Large Language Models (LLMs) to interact with external environments. However, current methods often struggle to handle massive and noisy candidate tools in long-context tool-calling tasks, limiting their real-world application. To this end, we propose Tool-DC, a Divide-and-Conquer framework for boosting tool-calling performance of LLMs. The core of Tool-DC is to reduce the reasoning difficulty and make full use of self-reflection ability of LLMs via a"Try-Check-Retry"paradigm. Specifically, Tool-DC involves two variants: 1) the training-free Tool-DC (TF), which is plug-and-play and flexible; 2) the training-based Tool-DC (TB), which is more inference-efficient. Extensive experiments show that both Tool-DC methods outperform their counterparts by a clear margin. Tool-DC (TF) brings up to +25.10% average gains against the baseline on BFCL and ACEBench benchmarks, while Tool-DC (TB) enables Qwen2.5-7B to achieve comparable or even better performance than proprietary LLMs, e.g., OpenAI o3 and Claude-Haiku-4.5.
