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 · pageggmx0225_5.png

What is the last underlined subheading on the page?

OCR text of the page · 1,895 characters; the scanned image itself is not published
SCOPE & FEASIBILITY
(INPUT TO GATE 2)
THIS PAGE IS REQUIRED IF QUALITATIVE RESEARCH IS CONDUCTED
TYPE OF RESEARCH CONDUCTED:
0000 * 0
Internal Focus Groups
Consumer Focus Groups
Consumer Panel
Online Qualitative
Online Quantitative
OTHER (DESCRIBE)
PRODUCT CONCEPT TESTED:
Explain how the concept was presented to those tested (visual attachments can be included):
Consumer workshops with approx 8 respondents and duration of 2 12 hours per group Recruit 12
to get 8 respondents
Age: 21-29 years of adult smokers; (50% between 21-25 years old)
Ethnic: 50% Caucasians, 50% mix of others
Education: all levels
Personal income: Annual salary of $20 000+
Living: Preference for adult smokers who live in their own flat/house
Working: All respondents work at least 20 hours a week. (Except 3 non working students per
group)
Attitude: all respondents with good verbal skills and a pleasure oriented lifestyle
Evaluation of the impact on the KOOL brand perception if the current packaging design would be
changed. Inputs to the brand communication on the pack: 00s, KOOL.
1 focus groups (female), with regular adult smokers of menthol brands (stand alones and line
extension menthol brands
50% adult Newport smokers
50% adult smokers of Marlboro menthol
1 focus groups (male), with regular adult smokers of Newport
2 focus groups (1 female, 1 male) with adult occasional menthol smokers
50% adult Newport smokers
50% adult smokers of Marlboro menthol
2 focus groups (mixed gender) with adult KOOL smokers
1 group with regular adult KOOL smokers
1 group with adult occasional KOOL smokers (regular brand: normal tobacco
LEARNING RESULTS:
Every effort should be made to gather these pieces of information, so a consistent evaluation occurs across
projects:
Mean Rating (10-pt scale)
Appeal
Importance
Concept
- Product
- Packaging
5
54611 2761
Source: https://www.industrydocuments.ucsf.edu/docs/ggmx0225
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
Learning ResultsLearning results
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

Mean Rating (10-pt scale)

25 characters9 tokens

Aux 2015

Wrong

The last underlined subheading on the page is:

"PRODUCT CONCEPT TESTED: Explain how the concept was presented to those tested (visual attachments can be included): Consumer workshops with approx 8 respondents and duration of 2 12 hours per group Recruit 12 to get 8 respondents"

279 characters64 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

Product

Focus Groups

Attitude

Evaluation

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

Marketing

280 characters64 tokens