datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Mind2Web-HTML-cleaned-lite-with-desc_w_taotrustsight-mind2web-injection
TrustSight Mind2Web-Injection Suite: Visual Grounding & Web Guardrail Benchmark
1. Tổng Quan Về Bộ Dữ Liệu (Dataset Overview)
TrustSight Mind2Web-Injection Suite là tập dữ liệu quy chuẩn (canonical tasks) tích hợp từ nghiên cứu Mind2Web-Injection (Lee & Choi, Tynapse Seoul, ArXiv 2026) vào hệ thống phòng thủ TrustSight.
Bộ dữ liệu tập trung đánh giá khả năng định vị không gian (Visual Grounding Localization) và chốt chặn an toàn (Capability Brokerage Guardrail)… See the full description on the dataset page: https://huggingface.co/datasets/tuandunghcmut/trustsight-mind2web-injection.Mind2Web-HTML-cleaned-lite-with-desc_w_tao_value_rationaleonline-mind2web-sortedmind2web-subset-human
Mind2Web Subset - Human Demonstrations
A collection of human-demonstrated web navigation tasks with detailed interaction traces. This dataset captures real browser interactions including clicks, typing, scrolling, DOM states, screenshots, and HTTP requests for web agent training and evaluation.
Overview
This dataset contains tasks performed by humans in real web environments, capturing:
Golden trajectories: Step-by-step sequences of actions (clicks, typing, navigation)… See the full description on the dataset page: https://huggingface.co/datasets/josancamon/mind2web-subset-human.ff-main-ec-online_mind2web_default-gc-chat_claude-mc-hf_qwen3_4b_inst_2507-lc-claude_age-r0ff-main-ec-online_mind2web_cc-gc-chat_claude-mc-hf_qwen3_4b_inst_2507-lc-claude_agent_so-r0