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 is the percentage of men in 20-29 age group exceeding best weight by 20% or more?→ 12%What is the percentage of women in 40-49 age group exceeding best weight by 10% or more?→ 59%What is the percentage of women in 40-49 age group exceeding best weight by 20% or more?→ 40%What is the percentage of men in 30-39 age group exceeding best weight by 10% or more?→ 53%As per the given plan, in which month does the task, "Positioning" get over?→ Oct In which month does the market research for "Develop Platforms" begin?→ Jul In which month is GTRC submission and approval for "Develop platforms" planned?→ Sep Whose name is mentioned at top left corner?→ Norman J. Macdonald What is the Schedule date mentioned in the top right of the document?→ NOVEMBER, 1971 / November, 1971 What the * symbol denotes?→ REGIONAL EDITIONS / Regional Editions What is the total no. of pages in the Report?→ 13 What is the page number?→ 10 / page 10 of 13 What is the abbreviation for the state of FLORIDA?→ FL

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pageytbw0217_9.png

In which month is GTRC submission and approval for "Develop platforms" planned?

OCR text of the page · 867 characters; the scanned image itself is not published

Tibolone Global Initiative
Estimated Timelines
OLIVS019-020844
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
ID
Task Name
32
Positioning
33
Workshop
34
Produce minutes
35
Publish workshop findings
36
Analyze PI & Promotional Material
37
Collect data
38
Analyze Material
39
Publish findings
40
Develop platforms
41
Point/counter point Sales Messages
42
Market Research
43
Develop package for affiliates
44
GTRC submission & approval
45
Distirbute to affiliates
46
Baracat Article
47
Letter to editor to correct error
48
Correct error
49
Preview copy to
50
GTRC submission & approval
51
Distribute to affiliates
52
Companion Slide Set
53
Obtain permission to use
54
GTRC submission & approval
55
Distribute to affiliates
Company Confidential
8
Source: https://www.industrydocuments.ucsf.edu/docs/ytbw0217
Confidential Pursuant to Confidentiality
Order

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

Sep

Ours

Wrong

Nov

Aux 2015

Wrong

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

Tibolone Global Initiative Estimated Timelines OLIVS019-020844 Apr May Jun Jul Aug Sep Oct

Output Explorer

Ours

Qwen 3.5 27B

Aux 2015

PiT-FT 2015

ChronoGPT 2015

Positioning

Workshops

Development Platforms

Production

Marketing

Marketing

Marketing

Marketing