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

what was the Ph.D Thesis on→ Nested Grids in Numerical Weather Prediction and an Application to a mesoscale Jet Streak. / Nested Grids in Numerical weather prediction and an application to a mesoscale Jet streak what is the Master's thesis→ Objective Mesoscale Temperature Forecasts.What is plotted along the x axis ?→ Year of birth What is plotted along the y axis ?→ age at menarche / Age at menarche what is the heading of the document ?→ pre-emergence chemical rates for sugar beets What is handwritten within brackets?→ days / Days What is the date mentioned at the bottom?→ 3-22-76 What is the numbers processed for Atlanta?→ 467 What is the numbers processed for Baltimore?→ 523 What type of data is it?→ preliminary data / Preliminary data What type of information should EPA use to identify conditions of use for inclusion in scoping?→ High Quality, Best Available Information / High quality, best available information / High quality, best available What is the page number?→ 9 / 9/Page / 9/page In tort cases what is the limit of liability?→ "reasonable foreseeablity"

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagekxvh0227_2.png

What is the date mentioned at the bottom?

OCR text of the page · 829 characters; the scanned image itself is not published

HYPERTENSION DETECTION AND FOLLOW-UP PROGRAM
PROCESSING TIME FOR HP06 FROM DATE OF FORM
TO DATE FORM RECEIVED
( DAYS)
Number
Shortest Longest 1
Mean
Standard
Center
Processed
Interval
Interval
Time
Deviation
Atlanta
467
11
126
24.9
15.1
Baltimore
523
6
75
26.9
7.2
Birmingham
335
282
20.8
19.4
Boston
434
34
15.0
5.5
Chicago
364
174
35.4
29. 1
Davis
213
7
71
19.0
6. 7
E. Lansing
291
10
75
19.8
8 . 2
Evans Co.
371
5
134
14. 1
12 . 3
Georgetown
2
311
10
128
27.5
14.8
Jackson
467
8
85
20.0
9.0
Los Angeles
504
12
93
28.5
15.6
Minneapolis
186
137
10.5
9.8
New York
534
117
27 . 3
15.5
Salt Lake
306
37
12.5
3. 3
Total
5306
282
22. 6
15.5
Includes forms requested from centers which have been identified as missing.
2 Includes all of Georgetown.
PRELIMINARY DATA
3-22-76
Source: https://www.industrydocuments.ucsf.edu/docs/kxvh0227

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

3-22-76

Ours

Exact match

3-22-76

Aux 2015

Wrong

22.6

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

HYPERTENSION DETECTION AND FOLLOW-UP PROGRAM PROCESSING TIME FOR HP06 FROM DATE OF FORM TO DATE FORM RECEIVED ( DAYS) Number Shortest Longest 1 Mean Standard Center Processed Interval Time