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

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagefxbn0020_3.png

what is the time period?

OCR text of the page · 741 characters; the scanned image itself is not published
201
BAN210IN
CPN/VPR/ BUYDOWN TRACKING/ANALYSIS
06/24/98 14:43
PROD BANLIB
GEOGRAPHIC OVERVIEW
-=== SELECTION CRITERIA -===
TIME PERIOD: 98
01/03/98 thru 06/12/98
BRAND: SS
TOTAL SAVINGS
GEOGRAPHY: XXXX
NATIONAL
--GEOGRAPHY--
& PV AFV/CTN ---TOT FV--- -SALES COST- -TOT ALLOC-
-REM ALLOC-
NATIONAL
76.1
2.93
169, 811, 466 170, 182, 698
1 NORTH EAST
72.1
2.88
32, 284, 513
32, 374, 260
2 SOUTHERN SA 73.5
3.03
70, 062, 053
70, 149, 354
3 US MILITARY
5 WESTERN SAL
71.2
2. 74
21, 179, 843
21, 250, 776
6 MID-WEST SA
86.9
2.90
46, 285, 076
46, 408, 325
7XXX NATIONAL
Enter-PF1---PF2---PF3---PF4---PF5---PF6---PF7---PF8---PF9---PF10--PF11--PF12---
HELP CRITE
LVLDN NAV
EXIT
52441 7467
Source: https://www.industrydocuments.ucsf.edu/docs/fxbn0020
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
981/03/98 thru 06/12/98
Models
4 of 4 columns · click a model to add or remove it

Ours

Partial · ANLS 0.95

01/03/98 thru 06/12/98

22 characters19 tokens

Aux 2015

Wrong

201

3 characters4 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong
63 characters64 tokens