DIVE

DIVE is a dataset that contains diverse, real-world tool-use tasks designed to evaluate the generalization capabilities of tool-using language models across various tool types and usage patterns.

Papers using DIVE (2)

DIVE-to-design: how a multi-agent workflow converts figure-centric literature into an ai-native hydrogen storage discovery engine2026 · 4 citesDIVE: Scaling Diversity in Agentic Task Synthesis for Generalizable Tool Use2026