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 · pagefhwc0228_12.png

What is ‘table 2’ ?

OCR text of the page · 864 characters; the scanned image itself is not published
Source: 1967/68 Menu Census
TABLE 2
FREQUENCY DISTRIBUTION OF SHEETNER USAGE
- Among Dieters and Weight Watchers
Times Used:
Persons Using
-- --Percent Total Uses---.
No.
%
Liquid
Granule
Tablet
All Other
Total
1 - 5 times
382
42.9%
48.9%
10.4%
26.9%
13.8%
100.0:
6 - 10 times
128
14. 4
52.2
8.5
29.5
9 .8
100.0%
11 - 20 times
225
25.2
46.7
7.3
38.1
7.9
100. C.
21 - 30 times
97
10.9
40.1
13.3
39.9
6.7
100. CT.
31 or more times
59
6.6
39.9
7.8
45.7
6.6
100.03
89 1
100 . 0%
- Among Non - Dieters -
Times Used:
Persons Using
- -Percent Total Uses---
No.
%%
Liquid
Granule
Tablet
All Other
Total
1 - 5 times
273
56.9%
43.8%
9.7%
34. 1%
12.4
100. C
6 - 10 times
78
16.3
50.5
9 . 7
31.6
8. 2
100 . 07:
11 - 20 times
86
17.9
23.2
13.4
55.0
8. 4
100.0:
21 - 30 times
30
6.3
28.8
3.3
50.6
17.3
100 . 0::
31 or more times
13
2.7
2.2
7.0
75.0
15.8
100.G.
480
100.0%
du/doc
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
frequency distribution of sweetner usageFrequency distribution of sweetner usage
Models
4 of 4 columns · click a model to add or remove it

Ours

Partial · ANLS 0.53

FREQGENCYOFHEETNERUSAGE

23 characters8 tokens

Aux 2015

Wrong

We are given a table that lists the frequency distribution of sheetner usage (liquid granules and tablets) for dieters and non-dieters. We are to answer the question: "What is 'table 2'?"

Looking at the table, we see that the total number of times sheetner is used

269 characters64 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters1 tokens

ChronoGPT 2015

Wrong

Table 2 Frequency Distribution of Sheller Usage

  • Among Dieters and Weight Watchers Times Used: Persons Using
    • Percent Total Uses--- No. % Liquid Granule Tablet All Other Total 1 - 5 times 382 42.9
200 characters64 tokens