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 · pageszpg0227_9.png

In Baltimore, what is the no. of referred cases, whose living status is unknown?

OCR text of the page · 843 characters; the scanned image itself is not published
TABLE 1
FIRST YEAR FOLLOW-UP STATUS AS OF 3/22/76
FOR STEPPED AND REFERRED CARE
Living Status Known
Living Status Unknown
Stepped
Referred
Stepped
Referred
N
%
N
N
ATLANTA
560
98.9
544
96.1
6
1. 1
22
3.
BALTIMORE
482
98. 4
488
99. 6
8
1. 6
2
0. 4
BIRMINGHAM
365
99 . 7
361
98. 9
0. 3
1 . 1
BOSTON
368
100.0
371
98. 4
O
0.0
1. 6
CHICAGO
255
100.0
259
99. 2
O
0.0
N
0.8
DAVIS
294
97 . 7
295
99. 0
2. 3
W
1. C
E. LANSING
291
99. 7
290
98. 3
0. 3
1.
EVANS CO.
343
99. 4
340
99 . 7
N
0. 6
0.3
GEORGETOWN
150
95.5
151
96. 2
4. 5
3.8
JACKSON
450
100.0
442
100 . 0
O
0.0
O
0.0
LOS ANGELES
441
100.0
430
100. 0
0
0.0
O
0.0
MINNEAPOLIS 426
100.0
428
100. 0
O
0. 0
O
0.0
NEW YORK
468
99. 4
459
99. 8
W
0. 6
0. 2
SALT LAKE
553
99. 3
535
98.2
0. 7
10
1. 8
TOTAL
5446
99. 3
5393
98.9
39
0. 7
62
1. 1
Source: https://www.industrydocuments.ucsf.edu/docs/szpgo
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
2
Models
4 of 4 columns · click a model to add or remove it

Ours

Exact match

2

1 characters2 tokens

Aux 2015

Wrong

To determine the number of referred cases whose living status is unknown, we first identify the number of cases where the living status is known (Living Status Known) and the number of cases where the living status is unknown (Living Status Unknown).

From the table, we see:

  • Living Status Known: Atlanta (5
309 characters64 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

Living Status Known Living Status Unknown Stepped Referred Stepped Referred N

Input:

165 characters64 tokens