Output Explorer

Every prompt in the paper, and what each model wrote back.

Answer a question about a scanned page from its OCR text. Scored by exact match and ANLS against the accepted answers.

13 of 5,330 prompts

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pageshlk0226_1.png

What is written in the page?

OCR text of the page · 69 characters; the scanned image itself is not published
ALABAMA
Source: http's://www.industrydocuments.ucsf.edu/docs/shik0226
System prompt · identical for every setup
Answer the question using only the OCR text from a single document page. Return only the answer, with no explanation. Preserve the answer wording from the OCR text when possible.
Expected answer
ALABAMAAlabama
Models
4 of 4 columns · click a model to add or remove it

PiT-FT 2015

Wrong

Empty response.

0 characters1 tokens

ChronoGPT 2015

Exact match

ALABAMA

7 characters3 tokens