Mind2Web
First page
Paper summary
A dataset for evaluating generalist web agents with 2,350 tasks across 137 websites and 31 domains.
Ask this paper
01
Broad web coverage: 137 real-world websites across 31 domains (travel, shopping, information seeking) - far more diverse than prior web benchmarks.
02
Generalization-focused: Tests cross-task, cross-website, and cross-domain generalization rather than in-distribution performance.
03
Realistic tasks: Uses real user tasks rather than synthetic scripts, capturing the messiness of actual web interactions.
04
Web-agent benchmark: Became a central benchmark for the 2024 explosion of web agents (WebAgent, WebVoyager, Browser Use, Operator).