AI Search Citation & Hallucination Auditor

Verify grounding fidelity, map claim tokens & catch phantom references

Fidelity Status
MODERATE_HALLUCINATION_RISK

Grounding Index Risk

Grounding Precision
75% 3/4 claims
Claim Grounding
3 grounded
1 unverified

Sentence-level claim verification

Phantom Citations
[4] 1 invalid ref

Tags absent or contradicted in source

0.45

Fuse.js fuzzy match threshold

Parsed AI Response Claims

Query: What is the quantum yield of GFP?

Sentence Level

Source Context & Overlap Evidence

Matched source passages and citation verification

Reference Database

Audit Summary & Proof Verification

Audit Slug ai-search-citation-auditor-74
Grounding Precision 75%
Grounded Claims 3 of 4
Phantom References [4]

Audit complete. Claim [4] exhibits unsupported assertions regarding ATP/flavins requiring exogenous cofactors, flagged as MODERATE_HALLUCINATION_RISK.

Grounding Fidelity and Phantom Citation Verification

Read the explanation

The auditor splits synthesized search responses into discrete claims, cross-referencing each claim against provided source passage texts. Verification combines sixty percent n-gram token overlap with forty percent fuzzy matching to score grounding confidence. When a claim references a missing tag like citation four, or drops below the twenty-five percent threshold, the audit flags a phantom reference. Selecting a preset or adjusting the sensitivity slider instantly recalculates grounding precision and updates verified claim tallies.

Super generates helpful tools and automates fact-checking across the internet proactively. If you enjoyed this tool, build your own with Super and share it with a friend.