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 · pagehxjw0081_4.png

In the first line, what is the first check box for?

OCR text of the page · 1,397 characters; the scanned image itself is not published
O NEW BRAND
D NEW ACCOUNT
O LINE EXTENSION
Forsyth/Field Sales Contact Person:.
Phone: .
Voice Mail:
Customer Services Contact Person:
Phone:
Completed by Forsyth)
(Completed by Forsyth)
Brand Name:
(Completed by Forsyth)
Brand Styles indicated below are those currently manufactured - if style(s) not needed, mark out; if additional styles
needed, check appropriate block (current styles will be filled out by Forsyth when Brand is determined).
O FF 85 O LT 85
O LT M 85 0 ULT 85 0 FF M 85
O FF BOX ( FF M BOX DJ LT 100 BOX
FF 100 _ LT 100 0 LT M 100 0 ULT 100 1 FF M 100 0 LT BOX LI LT MBOX DJ NON-FILTER
Do you expect this new brand to behave like any of the following brands?
O GPC O BASIC [ MONARCH O BEST VALUE
[Other)
Is the new brand replacing an existing brand?
(brand being replaced)
Please note any other pertinent information that may aid in the forecasting process (e.9., comparable accounts,
estimated share, pricing issues, etc.)
Are introductory promotions expected? O YES _ NO if yes, attach promotions form from Forsyth Planning Guide.
TO BE FILLED OUT BY FORSYTH/CUSTOMER SERVICES
. Check here if a forecast is needed for all accounts currently purchasing the brand (for line extensions only), or provide
. . ..:
account information in the spaces below.
. .. ..
...
When providing AIM data, Customer Services will pull 13 weeks of current information, when available.
51981 1752
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
NEW BRANDNew Brand
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

D NEW ACCOUNT

13 characters4 tokens

Aux 2015

Wrong

51981 1752

10 characters11 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

O FRSTH F LIT M F F O LIT M F O LIT M F O LIT M F O LIT M F O LIT M F O LIT M F O LIT M F O LIT M F O LIT M F O LIT M F O LIT M

127 characters64 tokens