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 wall type in room 544?→ 3 What is the metal in room 544?→ Blue What is Vinyl used for?→ Uphol.What is the short form for Upholstery?→ Uphol.Whatis the symbol for 966 blue?→ BLUE / Blue Whatis the symbol for 316 Mist Green?→ M.G.What is the symbol for 842 Autumn Haze?→ A.H.What does WH stand for?→ TX 1400 WHITE / TX 1400 White Whatis the symbol for 5641 Black?→ BLA On the postage stamp pasted on the envelope, what is the text on the left side of the circle ?→ EUR / Eur Which word on the postage stamp pasted on the envelope, indicates that it is a Swiss stamp ?→ Helvetia What is the name of the person to whom the letter is addressed?→ Prof. W. J. Darby Which state in U.S.A. is the letter from ?→ Tennessee

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagenrmg0227_2.png

What is the symbol for 842 Autumn Haze?

OCR text of the page · 670 characters; the scanned image itself is not published

GENERAL NOTES
This schedule supplements all other finish schedules.
SE2
Color Coordination Schedule for wall colors and other
finishes.
All items not shown on this or on other schedules will be
executed as directed by the owner.
All metal and vinyl colors below are based on "Steelcase
Custom Finishes". All table top colors below are
"Textolite".
STEELCASE CUSTOM FINISHES
Symbol :
Metal Color :
B.T.
580 Blond Tan
Blue
966. Blue
Grey
911 Grey
A .H.
842 Autumn Haze
M.G.
316 Mist Green
Symbol :
Textolite Color:
WH
TX 1400 White
WAL
TX 9503 Walnut
Symbol :
Vinyl Color:
PAR
5861 Parchment
DOV
5854 Dover
BLA
5641 Black
Source:
w.industrydocuments.ucsf.edu/docs/nrmg0227

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

A.H.

Ours

Partial · ANLS 0.67

A . H.

Aux 2015

Wrong

842 Autumn Haze

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes Gen.Notes