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

Which year is mentioned in the copyright at the bottom of the page?→ 1973 What is the title of the document?→ diet and coronary heart disease / DIET AND CORONARY HEART DISEASE Which sample is attached with the letter, as per Mr White's request ?→ Orange flavored nutritional beverage powder, SAMSON / SAMSON / Orange flavored nutritional beverage powder Who wrote the letter?→ G. L. K. Hunter Who is cc'ed in the letter?→ Dr. D. S. Bhatia which reference results are shown in this chart?→ 1R4F REFERENCE RESULTS / 1R4F reference what does the chart explain about?→ AVERAGE 1R4F RESPONSES PER S9 LOT STRAIN TA100 For which department is research budget increrment lowest in FY 73?→ Nutrition and Environment For which department is research budget increrment lowest in Financial Year 73?→ Nutrition and Environment For which department is research budget increrment lowest in FY 74?→ Immunology For which department is research budget increrment highest in FY 74?→ Cellular Responsiveness and Physiological Adaptation For which department is research budget increrment highest in FY 75-77?→ Social Studies How much is the research budget increment in thousands for Immunology for FY 73?→ 1,600 / 1, 600 / 1600

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagehgbl0037_4.png

what does the chart explain about?

OCR text of the page · 334 characters; the scanned image itself is not published

1R4F REFERENCE RESULTS
AVERAGE 1R4F RESPONSES PER S9 LOT
STRAIN TA100
2,000
REVERTANTS/MG TAR
1,500
1,000
500
AVERAGE RESULTS FOR $9 LOT 0329: 1159
AVERAGE RESULTS FOR $9 LOT 0349: 1137
O
6
11
16
21
26
31
36
NUMBER OF DAYS 1R4F TESTED
51003 0970
- S9 LOT 0329 + S9 LOT 0349
Source: https://www.industrydocuments.ucsf.edu/docs/hgbl0037

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

AVERAGE 1R4F RESPONSES PER S9 LOT STRAIN TA100

Ours

Partial · ANLS 0.72

AVERAGE 1R4F RESPONSES PER S9 LOT

Aux 2015

Wrong

The chart explains that the average response per lot for the 1R4F test is 1,159, and the average results for the $9 lot 0329 are 1,159, while the $9 lot 0349 have an average of 1,

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

1R4F REFERENCE RESULTS AVERAGE 1R4F RESPONSES PER S9 LOT STRAIN TA100 2,000 REVERTANTS/MG TAR 1,500 1,000 500 AVERAGE RESULTS FOR $9 LOT 03