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 · pagexpcf0004_1.png

What is the distribution quantity?

OCR text of the page · 1,146 characters; the scanned image itself is not published
Response Code Request Form
Fax # (910) 741-5327
Attn: Bonnie Tucker
RJR Brand:
Corporate
RJR Program #: 700399
Response Code: 286
Project Name; Target Dimensions June Names
Description: Names received from Target Dimensions for the month of June
Distribution:
Quantity
220,000
Distribution Vehicle: Other (Explanation in Comments)
Pct. Response
100.00
`# of Responders
220000
%
Age Verification:
DYES XNO
Timing:
DTS
06/30/97
Data Entry:
Expiration Date
06/30/97
Supplier None
Offer Complete
06/30/97
Days Offer Open
0
No. Days Tumaround
From 1 - 10)
Incoming mail N/A
Agency:
Agency Name
None
If M/A/R/C is supplier, do you
Contact Person
need a M/A/R/C P.O. Box?
N/A
Phone #
P.O. Box Title: N/A
Fulfillment:
Company
None
Job Number
Number of Items on Order Form
P.O. BOX .
City
State
Zip Code.
Comments: Mass Ad Names
Marketing Person Responsible for Project: _ Joyce Bagby
Production Contact:
Joyce Bagby
Response Code Requested By: Joyce Bagby
Copies of Confirmation Letter To:
Mary Cloutier
Joyce Bagby
Janet Myers
52266 1971
Was Bonnie Tucker
: RJRT Form 191 - WML 6/2 1/27
. . .
Source: https://www.industrydocuments:ucsf.edu/docs/xpcf0004
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
220,000
Models
4 of 4 columns · click a model to add or remove it

Ours

Exact match

220,000

7 characters8 tokens

Aux 2015

Wrong

Response Code Request Form Fax # (910) 741-5327 Attn: Bonnie Tucker RJR Brand: Corporate RJR Program #: 700399 Response Code: 286 Project Name; Target Dimensions June Names Description

184 characters64 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

Distribution Quantity

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

Distribution Vehicle

336 characters64 tokens