Supience Guide

Batch Information Check

Supience can examine multiple separate pieces of text in one session. Separate paragraphs with blank lines, then review each result individually and compare the overall pattern.

When batch checking helps

Batch checking is useful when reviewing several AI answers, claims, recommendations or short passages that belong to the same research task. It can help you notice which passages deserve closer verification.

What the results mean

Each result reflects observable signals involving evidence, specificity, uncertainty and verifiability. The result is not a truth score and does not establish that a claim is accurate or inaccurate.

How batch mode behaves

Each paragraph is analysed independently and shown as its own result. There is no combined summary, no average-quality verdict, and no single export of the whole batch.

Two practical limitations follow from this design:

If you need to revise a paragraph, replace its text in the textarea, leave the other paragraphs unchanged, and submit the whole batch again. The other paragraphs will be analysed again as part of the same submission.

Batch mode and file upload

If you upload a file, the extracted text is loaded into the textarea as a single block. Batch mode then splits it on blank lines as usual. This works when the source document already has clear paragraph separation.

Where a document uses a different layout — a wall of text with no blank lines, or a table that produces a single flattened block — the batch split may not match the document's structure. In that case, insert blank lines manually after extraction to separate the paragraphs you want analysed individually.

What to do next

For important claims, identify the exact statement, locate appropriate primary or authoritative sources, check dates and context, and compare the source evidence with the wording of the claim.

Tip: A group of high or low indicators does not replace checking the underlying evidence.

← Return to Supience Information Check