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 · pagefxyn0081_7.png

Which years were primarily the experimental ones?

OCR text of the page · 289 characters; the scanned image itself is not published
Background
Brand Tracker established in 1979:
= RJR had no standard tracking system in place
= Brand groups would field their own CAS/CCS
0 1979 - 1980 were primarily experimental years
Late 1980 a questionnaire design and data collection
methodology were established
Find Inder
51337 6211
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
1979 - 1980
Models
4 of 4 columns · click a model to add or remove it

Ours

Exact match

1979 - 1980

11 characters11 tokens

Aux 2015

Wrong

Based on the provided OCR text, the years are identified as:

  • 1979-1980: primarily experimental years
  • 1980: a questionnaire design and data collection methodology were established
  • 1981: Find Inder was established

Therefore, the years

240 characters64 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters1 tokens

ChronoGPT 2015

Wrong

0 1979 - 1980 were primarily experimental years Late 1980 a questionnaire design and data collection method were established

124 characters20 tokens