Test set

Test set B · CFPB disclosures, tables side by side

20 values on 2 pages, each checked by a person together with the place it is printed. Every page below shows a numbered bounding box on each checked value, then the answers.

Closing Disclosure, cash to close and transactions

Source: page 4 of 201403_cfpb_closing-disclosure_cover-H25B.pdf, CFPB sample form for a fictional loan (U.S. federal government work, public domain). The test reads it as a digital PDF page, with its text layer.

#	What it is	Section	Label	Value
1	Table cell	S1 · Calculating Cash to Close	row: Total Closing Costs (J) · column: Final	$9,712.10
2	Table cell	S1 · Calculating Cash to Close	row: Closing Costs Paid Before Closing · column: Final	– $29.80
3	Table cell	S1 · Calculating Cash to Close	row: Down Payment/Funds from Borrower · column: Did this change?	NO
4	Table cell	S1 · Calculating Cash to Close	row: Deposit · column: Loan Estimate	– $10,000.00
5	Table cell	S1 · Calculating Cash to Close	row: Cash to Close · column: Final	$14,147.26
6	Table cell	S2 · Summaries of Transactions: Borrower’s Transaction	BORROWER’S TRANSACTION · K. Due from Borrower at Closing	$189,762.30
7	Table cell	S2 · Summaries of Transactions: Borrower’s Transaction	BORROWER’S TRANSACTION · K. Due from Borrower at Closing · 03 Closing Costs Paid at Closing (J)	$9,682.30
8	Table cell	S3 · Summaries of Transactions: Seller’s Transaction	SELLER’S TRANSACTION · M. Due to Seller at Closing · 01 Sale Price of Property	$180,000.00
9	Table cell	S3 · Summaries of Transactions: Seller’s Transaction	SELLER’S TRANSACTION · M. Due to Seller at Closing · Adjustments for Items Paid by Seller in Advance · 12 HOA Dues 4/15/13 to 4/30/13	$80.00
10	Table cell	S3 · Summaries of Transactions: Seller’s Transaction	SELLER’S TRANSACTION · N. Due from Seller at Closing · 04 Payoff of First Mortgage Loan	$100,000.00

Loan Estimate, cost details

Source: page 3 of 201403_cfpb_loan-estimate_fixed-rate-loan-sample-H24B.pdf, CFPB sample form for a fictional loan (U.S. federal government work, public domain). The test reads it as an image of the page, with no text layer.

#	What it is	Section	Label	Value
1	Table cell	S1 · Loan Costs	Loan Costs · A. Origination Charges	$1,802
2	Table cell	S1 · Loan Costs	Loan Costs · A. Origination Charges · .25 % of Loan Amount (Points)	$405
3	Table cell	S1 · Loan Costs	Loan Costs · B. Services You Cannot Shop For · Appraisal Fee	$405
4	Table cell	S1 · Loan Costs	Loan Costs · C. Services You Can Shop For · Title – Lender’s Title Policy	$535
5	Table cell	S1 · Loan Costs	Loan Costs · D. TOTAL LOAN COSTS (A + B + C)	$5,672
6	Table cell	S2 · Other Costs	Other Costs · F. Prepaids · Prepaid Interest ( $17.44 per day for 15 days @ 3.875%)	$262
7	Table cell	S2 · Other Costs	Other Costs · G. Initial Escrow Payment at Closing · Property Taxes $105.30 per month for 2 mo.	$211
8	Table cell	S3 · Calculating Cash to Close	Calculating Cash to Close · Total Closing Costs (J)	$8,054
9	Table cell	S3 · Calculating Cash to Close	Calculating Cash to Close · Deposit	– $10,000
10	Table cell	S3 · Calculating Cash to Close	Calculating Cash to Close · Estimated Cash to Close	$16,054

Glossary

Term	Definition
OCR	Optical character recognition: reading the text, tables and marks of a page from its image or PDF.
engine	An OCR product or model that reads a page and returns its text, such as Mistral OCR, Landing AI or Reducto.
block	One part of a page as an engine returns it: a paragraph, a table, a heading or a figure.
bounding box	The rectangle that marks where something is printed on the page.
table cell	One value of a table, where a row and a column meet.
form field	A label on a form and the value written for it.
grounding	Linking each value of a reading to its bounding box on the page.
ground truth	The correct answers for a page, one fact per value, checkbox, choice or signature. Every reading is scored against them.
test set	The pages a scoreboard is measured on, with their ground truth.
grader	The program or model that compares a reading with the ground truth and counts what is right.