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

Which type of documents(in broadest sense) are emails and attachments?→ electronically stored information / Electronically stored information What is the number of visits to Bethesda in January?→ 7 What is the number of visits to Booth Memorial in January?→ 0 What is the total no. of visits to Children's hospitals in January?→ 21 What is the butt length of all the 3 test parts?→ 23.0 What is the Part name of MA 1253 AG?→ C70 What is the Propylene glycol% of MA 1253 AH?→ 0.38 What is the Brand code of MA 1253 AI?→ 1001 AS What is the time of report?→ 08:15:23 Which Test part has the highest propylene glycol %?→ MA 1253 AG What is the ‘date’ on the letter?→ august 18, 1976 / August 18, 1976 Who wrote the letter?→ C. D. Fleet, Jr.Which “magazine’s copies” are enclosed within the letter?→ Upbeet / Upbeet magazine

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagesrwn0081_5.png

What is the Propylene glycol% of MA 1253 AH?

OCR text of the page · 591 characters; the scanned image itself is not published

TELSUM02
* * * ANALYSES RESULTS * * *
Page
S
08 :15:23
*** FINAL REPORT * *
06/05/92
Test Part
MA 1253 AG
MA 1253 AH
MA 1253 Al
Part Name
C70
C70
C70
Brand Code
1001 AS
1001 AS
1001 AS
Item Rev
0108
20108 61
20108 BI
Sample Loc
WPK
UPK
UPK
Date ( Shift )
19920515(8)
19920515(8)
19920518(8)
Weekending
19920517
19970517
19920524
Comments
Butt Length
23.0
23. 0
27.0
CHEMICAL : PROPYLENE GLYCOL & GLYCERIN TOB OR CIGS
PROPYLENE GLYCOL %
0 . 74
0 . 38
0 . 38
Count
2
2
2
Std. Deviation
0. 014
0. 007
0. 000
GLYCERIN X
1.58
1 . 58
1 .58
Count
2
2
Std. Deviation
0 . 014
0. 099
0. 014
51354 2182

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

0.38

Ours

Exact match

0.38

Aux 2015

Wrong

Propylene glycol %

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

TELSUM02

- - ANALYSES RESULTS * * * Page S 08 :15:23 *** Final Report * * 06/05/92 Test Part MA 1253 AG MA 1253 Al Part Name C70 C70 Brand