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 · pagenxnw0228_1.png

What is the total amount in dollars?

OCR text of the page · 248 characters; the scanned image itself is not published
KHALIL GIRGI KHALEEFY
ORIENTAL STORES
BLISS St
Beirut Lebanon
PHONE 227911
Mr.
23/ 5/ 966
QUANTITY
DISCRIPTION
PRICE
TOTAL
9:00 2 18100
2. Table cloll 100 at 700= 14100
, Table cloth coloned 117 -
12:00
5.04
3
shift y ( 2) Boxes:
$806
8.00
18 16.00
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
$76.0076.00
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

$806

4 characters5 tokens

Aux 2015

Wrong

9:00 2 18100 2. Table cloll 100 at 700= 14100 , Table cloth coloned 117 - 12:00 5.04 3 shift y ( 2

98 characters64 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

$1,000,000

Question:

What is the total amount in dollars?

$1,000,000

75 characters32 tokens