Check your citations
Paste an AI answer, article, research summary, or source list. Caus shows how many distinct stories the citations really represent — and which of them just repeat the same coverage.
It shows how your sources relate — not whether the claim is true. Paste the cited text itself; Caus compares wording and never fetches URLs.
Checks run in your request only. Caus does not store your claim or source text, and usage analytics record counts, not the text you paste.
How this works
- 1
Paste what you are checking
An AI answer, article, research summary, or a list of sources. Caus recognizes common citation patterns — it does not open webpages.
- 2
Add the source text
Caus compares the wording you paste, so each citation needs the passage relevant to what you are checking.
- 3
Read the citation report
See how many distinct stories the citations represent, and which repeat the same coverage.
Caus recognizes citation patterns it can read — reference lists, numbered references, URLs, and structured source blocks. It never opens a webpage or looks anything up.
Nothing pasted yet — paste above, or add a citation by hand.
Checks run in your request only. Caus does not store your claim or source text, and usage analytics record counts, not the text you paste.
Prefer to enter citations one at a time? Add them by hand instead of pasting.
Context for this check — optional
Included in your Citation report, but not used to group sources.
Paste what you are checking, or add a citation, to run the check.
Advanced
Optional controls: adjust the grouping sensitivity or load more examples. Most users can leave these alone.
Threshold and sensitivity
The public default threshold is 0.55. It controls how much overlap in wording is needed before two citations are grouped into one story.
Higher threshold → stricter grouping
Fewer citations group together. Only very high overlap in wording counts as repeated coverage. Reduces false merges but may leave some real repeats separate.
Lower threshold → more sensitive grouping
More citations may group together. Catches more potential repeats but risks grouping genuinely different sources into one story.
The public default is conservative: near-verbatim repeated coverage is likely to group, but paraphrased or lightly rewritten coverage that shares meaning without sharing wording may stay separate.
The threshold adjusts grouping sensitivity, not truth or accuracy.
More examples
Load a ready-made example to see a repeated-coverage check in seconds. You can edit everything after loading.
Synthetic examples use fictional companies and publishers to avoid factual or legal risk — they make no claim about any real entity. They run the same conservative, production-safe check as your own sources.
Sample citation report — a synthetic example, computed with the same public, conservative check: 6 citations → 3 distinct stories.
Why grouped: These citations contain strongly overlapping coverage, so Caus grouped them as one story.
Why no strong match: No strong textual match was found with the other citations. This does not prove the source is independent, and heavily reworded coverage may be missed.
Load an example or paste your own sources above to run a live check — no sign-in.
Synthetic example. A citation report explains this grouping decision — it is not a truth verdict, not human-reviewed validation, and not an accuracy receipt.
How to read your citation report
Each report explains how Caus grouped the cited sources for one check. Here is what each field means.
- Story
- A set of citations that repeat substantially similar coverage, counted as one. A single citation with no strong match is a story of one.
- Repeated coverage
- Two or more citations whose wording overlapped enough to cross the grouping threshold — they appear to repeat the same underlying coverage and count as one story.
- No strong match
- A citation whose wording did not closely overlap any other citation. No strong textual match was found. That does not prove the source is independent, and this conservative check may miss paraphrased or lightly rewritten coverage.
- Strongest similarity
- The highest pairwise overlap score between any two citations inside a story. Only present for repeated-coverage stories.
- Average similarity
- The mean pairwise overlap score across all citation pairs inside a story. Gives a sense of overall overlap density.
- Threshold
- The cutoff used for this check. Citation pairs with overlap at or above this value are grouped into one story.
- Why grouped / Why no strong match
- A plain-English explanation of the grouping decision for each story — what the check found, not whether the claim is true.
Caveats: The report explains this result — it does not judge whether the claim is true. It is not human-reviewed validation and not an accuracy receipt. No strong match is not a certificate of independence: this conservative check may miss paraphrased or lightly rewritten coverage that shares meaning but not wording.
Caus shows how your cited sources relate — how many distinct stories they represent, and which repeat the same coverage. It does not decide whether the claim is true.