EN-Web-Understanding

English web screen understanding data used to train the KORMo-Web agents. Each subset is one task; pick it by name.

Text only. Screenshots are not included; image_files gives the file name of the screen(s) each row was made from.

from datasets import load_dataset
ds = load_dataset("KORMo-VL/en-web-understanding", "grounding", split="train")

Subset	Rows	Task
page_caption	272,055	Grounded description of the whole screen (element names with 0–1000 coordinates)
screen_qa	1,219,623	Question answering about what is on the screen
ocr_read_text	876,419	Read the text in a given region or the whole screen
grounding	1,547,373	text_to_point / point_to_text / appearance_to_point (0–1000 coordinates)
link_destination_choice	134,918	Choose where a link leads
screen_change_caption	328,404	Describe what changed between two screens and how
action_between_screens	392,935	Which action (click / scroll) turns screen 1 into screen 2
action_between_screens_think	253,376	Same as above, with reasoning
next_action	576,721	Next action toward a goal (click)
next_action_think	174,101	Same as above, with reasoning
next_screen_prediction	210,343	Predict the screen after a click
next_screen_prediction_detailed	254,616	Predict the screen after a click or scroll, in detail
Total	6,240,884

Fields

Field	Description
id	Unique row id
lang	en
task / subtask	Task name and variant (e.g. click, scroll, text_to_point)
question	User prompt
answer	Target answer. For action tasks, a JSON tool call such as {"name": "click", "arguments": {"x": 500, "y": 300}}
reasoning	Reasoning text when the task has one, else null
image_files	Screenshot file name(s) the row refers to

Coordinates are normalized to 0–1000.

Notes

• Session, token and password values inside URLs were replaced with [MASKED].