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 · pagefxcv0079_4.png

What does the unshaded bar represent?

OCR text of the page · 266 characters; the scanned image itself is not published
14 DAY SMOKE INHALATION STUDY IN MICE TAD-ATS-020
TIDAL VOLUME MEANS AT PRE-EXPOSURE AND 60 MINUTE
EXPOSURE PERIOD AVERAGE
BOTH SEXES COMBINED
0.2T
0 . 15+
TIDAL
VOLUME 0. 1+
mls
0 . 05+
MC
TLE
TME
THE
ALE
RME
RHE
EXPOSURE GROUPS
PRE-EXP
WITHIN 60 MIN AVE
51493 8410
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
Pre-ExpPre-expPre-exposure
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

WITHIN 60 MIN AVE

17 characters9 tokens

Aux 2015

Wrong

TIDAL VOLUME MEANS AT PRE-EXPOSURE AND 60 MINUTE EXPOSURE PERIOD AVERAGE BOTH SEXES COMBINED 0.2T 0 . 15+ TIDAL VOLUME 0. 1+ mls

129 characters64 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

14 DAY SMOKE INHALATION STUDY IN MICE TAD-ATS-020 TIDAL VOLUME MEANS AT PRE-EXPOSURE AND 60 MINUTE EXPOSURE PERIOD AVERAGE BOTH SEXES COMBINED 0.2T

147 characters64 tokens