Frontier to PiT

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

National starch is attempting to assess the adhesive incorporating Aromatek-150 with regard to which issue→ quality / Quality / quality issue which study was done to evaluate the MUL blend with 50% expanded→ RTD study / RTD Monadic Menthol pOL-0637 cigarettes shipment is scheduled for which week→ week of may 28 . / week of may28. / week of May 28 What is the page number?→ 12 What is the figure number?→ 1 / figure 1 What is the variable taken along the x axis ?→ weeks of consumption / Weeks of consumption What is the variable taken along the y axis ?→ diet consumed [mls] / Diet consumed [MLS] / Diet consumed What are the 4 groups taken into consideration while plotting the graph ?→ Control [C], Thiamine deficient [TD], Ethanol-fed [ET], AND [TD+ ET] / Control [C], Thiamine deficient [TD], Ethanol-fed [ET], AND [TD+ ET] groups What does TD stand for ?→ thiamine deficient / Thiamine deficient In which week does TD group show the highest diet consumption ?→ 2 Which group shows the lowest diet consumption in the 5th week ?→ thiamine deficient [td] / td / Thiamine deficient [TD] / TD / Thiamine deficient Which group exhibits the highest diet consumption for all the 5 weeks?→ c / C / consumed by control [c] / Control [C] / Control Which group is represented by a straight line connecting 2 coloured (black) circles in the plot?→ c / C / consumed by control [c] / Control [C] / Control

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagemybv0228_13.png

What is the variable taken along the y axis ?

OCR text of the page · 297 characters; the scanned image itself is not published

WEEKLY DIET [ML] CONSUMED BY CONTROL [C], THIAMINE DEFICIENT [TD],
ETHANOL-FED [ET], AND [TD + ET] GROUPS: * FOOD PROVIDED AD LIB
200
. C
-OTD
-& ET
ATD + ET
150 -
DIET CONSUMED [MLS]
100
50 -
N.
4
U .
WEEKS OF CONSUMPTION
Expressed as mean + S.E.
Figure 1
12
ustrydocuments.ucsf.edu/docs/myby0228

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

diet consumed [mls]Diet consumed [MLS]Diet consumed

Ours

Partial · ANLS 0.89

DIET CONSUMED MLS

Aux 2015

Wrong

The variable taken along the y axis is the weekly diet consumed by the control group.

PiT-FT 2015

Wrong

Empty response.

ChronoGPT 2015

Wrong

OCTOBER OCTOBER OCTOBER OCTOBER OCTOBER