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 · pagermdw0217_1.png

Which date is printed on the bottom right corner of the page?

OCR text of the page · 1,940 characters; the scanned image itself is not published
STATEGIC PUBLICATIONS DEVELOPMENT
TRACKING REPORT
MS No.
Article
Author
Journal
Status
(Focus)
MS #1
Recent Insights into the
M. Dey, C.R. Lyttle,
Maturitas
Published in April
Components)
Varying Activity of
J.H. Pickar
Supplement)
Estrogens
M. Rhoa to be sent
reprints (W-A
funded supplement?)
MS #2
Comparative Tolerability of | M.G. Curtis
Drug Safety
Invited manuscript
(SERMS)
the Selective Estrogen
in development
Receptor Modulators in
Breast Cancer Treatment
July 30 deadline by
and Prevention
journal
MS #3
Treatment of Urinary Tract
C. Maloney
Urology Nurses
Invited manuscript
(PVC)
Infections in
Online
submitted April 21,
Postmenopausal Women
2000
Resent to journal
6/13/00
MS #4
Hormone Replacement
L. Mosca
Current
Published
(CVD)
Therapy in the Prevention
Atherosclerosis
July 2000; 2(4):297-
and Treatment of
Reports
302
Atherosclerosis
MS #5
Hormone Replacement
D. Archer
Obstetrics and
Paper submitted
(Patien
Therapy: Effect of
J. Pickar
Gyncecology
1/4/00; Reviewers
Management)
Progestin Dose and Time
comments received
Since Menopause on
3/16/00; Revised
Endometrial Bleeding
paper submitted
5/18/00
MS #6
Breast Cancer and HRT in
L. Nachtigall
The Female Patient
Manuscript ready for
(Breast
Perspective
submission.
Cancer)
MS #7
Role of HRT in the
H. Fillit
Archives of
Manuscript in
(CNS)
Prevention of Alzheimer's
Neurology
development
Disease
MS #8
Estrogen Tissue Effects and | B.A. Eskin
Journal of Clinical
Author to send
(CVD)
Adrenergic
Endocrinology and
outline in early June;
Cardioprotection with
Metabolism
priority is low
Conjugated Equine
Estrogens
MS #9
Leisure World Cohort:
A. Paganini-Hill
TBD
Data analysis in
(B of B)
Update and Review
progress
Confidential
Page 1
July 5, 2000
DESIGNWRITE . 189 WALL STREET, PRINCETON, NEW JERSEY 08540 . 609/924-1116 . FAX: 609/924-6648
Confidential Pursuant to Confidentiality
DUROJ023-001868
Order
Source: https://www.industrydocuments.ucsf.edu/docs/rmdw0217
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
July 5, 2000
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

Order

5 characters2 tokens

Aux 2015

Wrong

The document contains a list of 10 journal articles, each with a specific publication date, author, journal, status, and a focus. The question asks for the date printed on the bottom right corner of the page.

Looking at the list, the bottom right corner of the page is typically the last line of

296 characters64 tokens

PiT-FT 2015

Wrong

<|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|>

120 characters64 tokens

ChronoGPT 2015

Wrong

Chronological Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date Date

163 characters64 tokens