What this category's data looks like
This page focuses on three types of materials in student awards, grants, loans, and hardship determination scenarios: the family economic hardship determination form, the scholarship/grant application form, and the loan receipt. The determination form comes from student submissions and class review, is updated by academic year or semester, and its fields include family members, income, hardship level, and a signature field; the application form opens with scholarship/grant project batches and includes student ID, project name, amount, date, and counselor signature; the loan receipt is generated by a bank or financial aid system and includes contract number, disbursement date, amount, and receipt stamp. When these three types of materials enter this feature, the determination form serves as the source of basic identity and hardship level, the application form serves as the source of facts for the current application, and the loan receipt serves as the source of fund arrival and contract performance, jointly supporting the verification and discrepancy localization of the same object's name, quantity, and date across two materials.
Configuration and implementation checks
This part is executed by application settings (response limit, etc.), which output the length of a single answer and the range of context that can be accommodated. The "discrepancy list" is then completed by the model. The response limit is controlled by a slider for the length of a single answer; if set too low, the answer will be truncated, and if set too high, consumption will increase.
| Object | Check | Verification action |
|---|---|---|
| Family economic hardship determination form | When compared with the application form and loan receipt, the student name, student ID, hardship level, family income, and signature date must be consistent | In application settings, check the response limit switch and slider position, then submit the two materials to let the model compare field by field and output a discrepancy list |
| Scholarship/grant application form | When compared with the determination form and loan receipt, the project name, application amount, application date, and counselor signature must be consistent | Check whether the end of the answer is complete, and verify whether the single-answer length corresponding to maxResponse matches the length of the discrepancy list |
| Loan receipt | When compared with the determination form and application form, the contract number, disbursement amount, disbursement date, and receipt stamp must be consistent | Record the number of entries in the citation array, and separately check the citation text tokens corresponding to quoteMaxToken; do not compare the two for equality |
| Context range | The total amount of determination form, application form, and loan receipt fields that can be accommodated in a single instance must cover the materials needed for verification | Check whether maxContext accommodates the key fields of the three materials, to avoid missing discrepancy fields due to insufficient range |
Three common errors
- Symptom: The name or student ID of the same student does not match between the determination form and the application form. Cause: The two materials are entered through different entry points, and the primary key alignment for the same object was not performed, resulting in missing items in the discrepancy list.
- Symptom: The disbursement amount and date on the loan receipt are inconsistent with the amount and date on the application form. Cause: The versions were not synchronized; after the receipt was updated, it was not written back to the ledger, and the old version was still referenced.
- Symptom: The determination form or application form lacks a counselor signature, or the loan receipt lacks a receipt stamp. Cause: The material circulation step did not verify the completeness of signatures and receipts, and they were treated as valid records when entering verification.
How to verify the setup
Feature acceptance: ask a question that requires a long answer, for example listing all field discrepancies across the three materials, and see whether the end of the answer is complete, confirming that the response limit did not cause the sentence to break in the middle.
Business test: use a real record to find fields inconsistent with the ledger, and hand it to a reviewer for confirmation and write-back.
- In application settings, check the response limit switch and slider position, and confirm that the single-answer length matches the length of the discrepancy list.
- Compare the usage of two runs to confirm that the change in single-run usage after adjusting the limit is within an acceptable range.