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 · pagetkgj0224_1.png

What is the age limit of the Screening Criteria?

OCR text of the page · 1,641 characters; the scanned image itself is not published
AW06- 04
bri
bellomy
research
April Claim Substantiation Study Recruit
Cost Estimate
February 17, 2006
The following will confirm details of the recruiting for the two-day Claim Substantiation Study:
Methodology:
o Recruit 200 consumers for 100 to show for a two-day study to be held at RJRT's R&D building.
75% or 150 of the recruits should be non-smokers
o 25% or 50 of the recruits should be smokers
o Other screening criteria includes:
21-49 years of age
o
Standard security screening
Must agree to smell a small amount of tobacco smoke
o
Cannot be currently participating in any other cigarette research
Attempt to recruit an even split of ages (21-34 & 35-49) and gender within smokers and
non-smokers
O
The study will be held on April 26" and 27" with a morning session (10:00-1 1:00 am) and an
afternoon session (2:00-3:00 pm)
Recruited consumers will attend one session on each of the two days. We will try to recruit
consumers to attend the same session time on each of the two days.
o
Gratuity of $75 will be paid for participating on both days
BRI Responsibilities:
Recruit 200 consumers for 100 to show that meet the criteria listed above with an even split
between the two session times available
Place reminder calls and send reminder letters with directions to the facility
o
Provide hostess for check-in and payment of gratuity on both days
Timing:
The following provides the estimated timing for this project:
Task
Timing
Project Approval
4/3/06
Develop Recruit Screener
4/3/06-4/6/06
Recruiting
4/7/06-4/23/06
Groups
April 26 & 27 at
10:00am & 2:00pm
53420 6775
Source: https://www.industrydocuments.ucsf.edu/docs/tkgj0224
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
21-49 years of age
Models
4 of 4 columns · click a model to add or remove it

Ours

Exact match

21-49 years of age

18 characters9 tokens

Aux 2015

Wrong

The age limit of the Screening Criteria is 21-49 years of age.

62 characters19 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

OCTOBER 26, 2006

OCTOBER 26, 2006

OCTOBER 26, 2006

OCTOBER 26, 2006

OCTOBER 26, 2006

OCTOBER 26, 2006

OCTOBER 26,

120 characters64 tokens