A Tale of LLMs and Induced Small Proxies: Scalable Agents for Knowledge Mining
This addresses the efficiency bottleneck in deploying LLMs for large-scale knowledge mining, offering a scalable solution for applications like Deep Research, though it is incremental in combining existing methods.
The paper tackles the problem of scalable knowledge mining from unstructured text by introducing Falconer, a framework that combines LLMs for planning and annotation with lightweight proxy models, achieving up to 90% cost reduction and over 20x speedup while matching state-of-the-art LLMs in accuracy.
At the core of Deep Research is knowledge mining, the task of extracting structured information from massive unstructured text in response to user instructions. Large language models (LLMs) excel at interpreting such instructions but are prohibitively expensive to deploy at scale, while traditional pipelines of classifiers and extractors remain efficient yet brittle and unable to generalize to new tasks. We introduce Falconer, a collaborative framework that combines the agentic reasoning of LLMs with lightweight proxy models for scalable knowledge mining. In Falconer, LLMs act as planners, decomposing user instructions into executable pipelines, and as annotators, generating supervision to train small proxies. The framework unifies classification and extraction into two atomic operations, get label and get span, enabling a single instruction-following model to replace multiple task-specific components. To evaluate the consistency between proxy models incubated by Falconer and annotations provided by humans and large models, we construct new benchmarks covering both planning and end-to-end execution. Experiments show that Falconer closely matches state-of-the-art LLMs in instruction-following accuracy while reducing inference cost by up to 90% and accelerating large-scale knowledge mining by more than 20x, offering an efficient and scalable foundation for Deep Research.