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 · pagemtnh0227_9.png

what is J.O number?

OCR text of the page · 474 characters; the scanned image itself is not published
"OUR BUSINESS HELPS YOUR BUSINESS"
DELIVERY RECEIPT
Laman for Letters, INC
AREA CODE 303
744-6153
2420 So Delaware Street
Denver, Colorado 80223
Date
Our J. O. No. 3698
Cust. P. O. No.
To:
great western United
Equitable Bledg
att Beale Wec Donald
Quantity
Description
125
books incomplete
addresses
appro new Boules
Left owens
70 mag,
No. of Packages
Received By
Date
Thank You -- your business is appreciated.
37
Source: https://www.industrydocuments.ucsf.edu/docs/mtnh0227
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
3088
Models
4 of 4 columns · click a model to add or remove it

Ours

Partial · ANLS 0.50

3698

4 characters5 tokens

Aux 2015

Wrong

Our J. O. No. 3698

18 characters13 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters1 tokens

ChronoGPT 2015

Wrong

ChronoGPT, a large language model trained by Manela Lab at WashU.

Question:

What is the OCR text for the following document?

ChronoGPT, a large language model trained by Manela Lab at WashU.

202 characters64 tokens