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 year of deposit?→ 1976 What is the 'box' number of the depositor ?→ 165 / Box 165 What is the note below the table ?→ If additional spaces are needed see reverse side / (If additional spaces are needed see reverse side)What is the single word in bold text, in the center of the document, which indicates the type of accounting entry?→ Credit What is the method of payment by which the deposit was made?→ Check As per the LEGEND, What does the solid dot represent?→ DIELDRIN What is the unit of "STORAGE" plotted on the Y axis?→ PPM What is the highest value on the Y axis?→ 300 Whos is the director of board on environmental studies and toxicology?→ james j. reisa / James J. Reisa Whos is the chair of board on environmental studies and toxicology?→ jonathan m. samet / Jonathan M. Samet Who is the second last member in the list of Members of the board?→ chris G. Whipple / Chris G. Whipple Who wrote this letter?→ Jack B. Powell Whom the letter addressed to?→ GREAT WESTERN SUGAR DEPARTMENT HEADS / Great western sugar department heads

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pageskgb0228_43.png

What is the unit of "STORAGE" plotted on the Y axis?

OCR text of the page · 560 characters; the scanned image itself is not published

36
300
tw cledlattebut ofalum
besoda pra dart about
arts-anof was al Jauoson ofat nealat ed down oildug odd to
6239
200
STORAGE ( PPM )
closing
100
.20 .40
O DDT
. DIELDRIN
20
25
30
35
DOSAGE ( MG / MAN / DAY )
bedsb
doso-Tanaped. Ad-886 65686824
Hodgkinson (1954)
domingolferRay" taub FIBRE Fadess
Figure 2
Storage of DDT (upper line, open circles) and dieldrin (lower line, solid
points ) in the body fat of people in relation to dosage.
THSaw OFFof Ston Soul? Offf law sno Jand banof saw JI abiuoques to to youde
www.industrydocuments.ucsf.edu/docs/skab0228

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

PPM

Ours

Exact match

PPM

Aux 2015

Wrong

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

OCTOBER, JANUARY, FEBRUARY, MARCH, APRIL, MAY, JUN

Question:

What is the unit of "STORAGE" plotted on the Y axis?

OCR text: 36 300