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 · pagekkwg0227_7.png

What is subheading J?

OCR text of the page · 847 characters; the scanned image itself is not published
I. COOPERATIVE RESEARCH PROJECTS:
1. With other Divisions of U. S. Army Medical Research and Nutrition Lab.
No. 3-11:
Human pyridoxine requirement and the interrelationship of
pyridoxine to the dietary macronutrients
No. 3-20:
Effect of autooxidation on L-ascorbic acid
2. With groups outside of U. S. Army Medical Research and Nutrition Lab.
J. ANALYTICAL SERVICE:
Routine analysis to support research projects within Chemistry Division
and other cooperating groups
K. MISCELLANEOUS STUDIES:
1. Nutritional studies with microcrystalline cellulose (Final)
2.
Possible relationship of magnesium to human kidney stone formation
(Preliminary progress )
3.
Survival and fall-out shelter studies
4. ICNND Nutrition Surveys
5. Effect of deoxypyridoxine on tryptophan metabolism in the human
7
Source: https://www.industrydocuments.ucsf.edu/docs/kkwg0227
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
ANALYTICAL SERVICEAnalytical Service
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

ANALYTICAL SERVICE: Routine analysis to support research projects within Chemistry Division and other cooperating groups

120 characters22 tokens

Aux 2015

Wrong

I. COOPERATIVE RESEARCH PROJECTS:

  1. With other Divisions of U. S. Army Medical Research and Nutrition Lab. No. 3-11: Human pyridoxine requirement and the interrelationship of pyridoxine to the dietary macronutrients No
219 characters64 tokens

PiT-FT 2015

Wrong

<|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|> <|<assistant|>

120 characters64 tokens

ChronoGPT 2015

Wrong

I. COOPERATIVE RESEARCH PROJECTS:

  1. With other Divisions of U. S. Army Medical Research and Nutrition Lab. No. 3-11: Human pyridoxine requirement and the interrelationship of pyridoxine to the dietary macron
208 characters64 tokens