MiniWoB++

MiniWoB++ is a benchmark dataset that contains a collection of web navigation tasks used to evaluate the performance of web agents in following natural language goals through multi-step browser interactions.

Papers using MiniWoB++ (3)

A Zero-shot Language Agent For Computer Control With Structured Reflection2023 · 5 citesWeasel: Out-of-Domain Generalization for Web Agents via Importance-Diversity Data Selection2026AdaPlanner: Adaptive Planning from Feedback with Language Models2023 · 13 cites