Lomekwi: Resource-Bounded Tool Discovery in LLM Agents
Abstract
Existing tool-use benchmarks report a single success rate for complex, multistep tasks. Inspired by ideas from cognitive science, we distinguish tool use from tool discovery and decompose the latter into curiosity (the model's ability to discover the parts needed to build the tool), recognition (the model's ability to discover the process of creating the tool), and efficiency (the model's use of the tool after creation). We show that this framework can be applied to existing discovery tasks, such as Voyager. In addition, we provide evidence that recognition inversely scales with model size, and we introduce and analyze a class of combinatorial games that demonstrates this. We further observe inverse scaling in a separate environment designed to emulate real-world tasks.
Cite
@article{arxiv.2607.16961,
title = {Lomekwi: Resource-Bounded Tool Discovery in LLM Agents},
author = {Roshan Klein-Seetharaman and Daniel Wang and Andrew Xu},
journal= {arXiv preprint arXiv:2607.16961},
year = {2026}
}
Comments
All authors contributed equally. 17 pages, 6 figures. Presented at the 2026 Conference on Learning Theory Workshop on Learning in an Agentic World