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 · pagerrcw0217_2.png

What is the page number of ‘introduction’ in the TABLE OF CONTENTS?

OCR text of the page · 1,091 characters; the scanned image itself is not published
ERT/HRT Expert Forums
TABLE OF CONTENTS
INTRODUCTION .
SEXUALITY CHANGES IN PERI- AND POSTMENOPAUSAL WOMEN..
. 5
Preliminary Discussion Guide...
Proposed Faculty ..........
......8
Value-Added Enduring Materials............
.....11
ACUTE COGNITIVE SYMPTOMS ASSOCIATED WITH
THE MENOPAUSAL TRANSITION ....."
12
Preliminary Discussion Guide......
......13
Proposed Faculty ...............
...15
Value-Added Enduring Materials .............
.... 18
MECHANISMS UNDERLYING THE EARLY CARDIOVASCULAR
EFFECTS OF ERT/HRT ........
........ 19
Preliminary Discussion Guide..
...20
Proposed Faculty .........................."
.21
Value-Added Enduring Materials ..............
......24
PREVENTING OSTEOPOROSIS: ESTABLISHING
THE ROLE OF ERT/HRT IN BONE HEALTH ...........
... 25
Preliminary Discussion Guide...............
26
Proposed Faculty........
.....27
Value-Added Enduring Materials.......................
....30
BUDGET ESTIMATE.............
DESIGNWRITE . 189 WALL STREET, PRINCETON, NEW JERSEY 08540 . 609/924-1116 . FAX: 609/924-6648
Confidential Pursuant to Confidentiality
Order
COBBP007-000016
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
3
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

. 5

3 characters4 tokens

Aux 2015

Wrong

26

2 characters3 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters1 tokens

ChronoGPT 2015

Wrong
### Introduction
### Introduction
### Introduction

The same 17-character fragment repeats 21 times until the token limit. Showing the first three.

360 characters64 tokens