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 · pagelngn0226_5.png

What does the numbers in the Y-axis indicate?

OCR text of the page · 254 characters; the scanned image itself is not published
Number of Patients Currently in Each Interval
250
212
217
200
Number of Patients
150
119
100
42
50
PreOp
6 mo
12 mo
24 mo
36 mo
48 mo
60 mo
PROTECTED DOCUMENT. DOCUMENT SUBJECT TOSPRITECHIVESORDERV. industrydocuments.ucsf.edu/docs/Ingn0226
DEPUY048356009
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
Number of Patients
Models
4 of 4 columns · click a model to add or remove it

Ours

Exact match

Number of Patients

18 characters4 tokens

Aux 2015

Wrong

250

3 characters4 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

Number of Patients Currently in Each Interval 250 212 217 200 Number of Patients 150 119 100 42 50 PreOp 6 mo 12 mo 24 mo 36 mo 48 mo 60 mo PROTECTED DOCUMENT. DOCUMENT SUB

172 characters64 tokens