Score: 2

Procedural Environment Generation for Tool-Use Agents

Published: May 21, 2025 | arXiv ID: 2506.11045v2

By: Michael Sullivan, Mareike Hartmann, Alexander Koller

Potential Business Impact:

Teaches AI to use tools better with fake practice.

Business Areas:
Simulation Software

Although the power of LLM tool-use agents has ignited a flurry of recent research in this area, the curation of tool-use training data remains an open problem$-$especially for online RL training. Existing approaches to synthetic tool-use data generation tend to be non-interactive, and/or non-compositional. We introduce RandomWorld, a pipeline for the procedural generation of interactive tools and compositional tool-use data. We show that models tuned via SFT and RL on synthetic RandomWorld data improve on a range of tool-use benchmarks, and set the new SoTA for two metrics on the NESTFUL dataset. Further experiments show that downstream performance scales with the amount of RandomWorld-generated training data, opening up the possibility of further improvement through the use of entirely synthetic data.

Country of Origin
🇩🇪 Germany


Page Count
19 pages

Category
Computer Science:
Machine Learning (CS)