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

To Whom is this addressed to?→ dr. darby / Dr.Darby What was the date on the slip?→ 5-5-86 What time is mentioned on the slip?→ 11:57 / 11:57A What is the phone number given in the "Phone" field?→ (916)752-2442 / (916) 752-2442 which country's tobacco brand tracker?→ US what is the rate of usual brand for roll-your-own ?→ 1.3%what is the rate of usual brand for factory-made?→ 97.9%what is the rate of no usual brand?→ 0.8%When is the Summary Of the Week?→ 9:00 - 12:00 When is the student panel?→ 1:00 - 4:00 What is the page number?→ 29 Which committee is this→ THE FOOD PROTECTION COMMITTEE What is the Program # ?→ 701998

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagepkkn0226_7.png

what is the rate of usual brand for factory-made?

OCR text of the page · 252 characters; the scanned image itself is not published

2012 US Tobacco Brand Tracker
Usual Brand
Factory-Made Versus Roll-Your-Own
Usual Brand
%
Factory-Made
97.9%
Roll-Your-Own
1.3%
No Usual Brand
0.8%
53717 2573
Base: Adult Smokers 21+ (21,294)
Source: https://www.industrydocuments.ucsf.edu/docs/pkkn0226

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

97.9%

Ours

Exact match

97.9%

Aux 2015

Wrong

2012 US Tobacco Brand Tracker Usual Brand Factory-Made Versus Roll-Your-Own Usual Brand % Factory-Made 97.9% Roll-Your-Own 1.3% No Usual Brand 0.8% 537

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

Factory-Made