Hallucination detection

Catch hallucinations
claim by claim.

A fluent answer can still be wrong in one number. Latent Lens splits every answer into single claims and checks each one against your source, so you see exactly which sentence is the problem.

The hallucinations it catches

  • Numeric drift: "60 days" in the answer when the source says 30.
  • Fabricated entities: names, teams or products that appear nowhere in the source.
  • Contradictions: a rule flipped from "non-refundable" to "refunded in full".
  • Unsupported claims: statements the source simply does not back.
  • Meaning errors: role reversals, scope changes and antonyms that share every word with the source, caught by the HHEM-2.1-Open model.

How it works

The response is split into atomic claims. Each claim is matched to the closest sentence in the ground truth and labeled supported, partial, contradicted or unsupported. Deterministic checks run first, so a wrong number is never excused by a model. The meaning check can only make a verdict stricter or rescue a reworded claim; it never overrules a deterministic catch.

Long documents are split into sentence-aligned windows, so a fact on page 30 is still found.

Without ground truth

If you do not paste a source, claims cannot be verified. The report then rates hallucination risk instead: specific figures, names and citations that nobody can check, and overconfident wording. The result is clearly marked as low confidence.

Honest limits

No automatic check is perfect. Temporal order ("before" vs "after") is a known gap of the open model; an LLM judge or a human expert review covers it. Try it now in the free checker or learn about RAG evaluation.

Questions

Frequently asked

What is an AI hallucination?

An AI hallucination is a statement in a model's answer that is not supported by, or contradicts, the source it should be based on, such as a wrong number, an invented name or a flipped rule.

How does Latent Lens detect hallucinations?

The answer is split into atomic claims, each claim is checked against your ground truth and labeled supported, partial, contradicted or unsupported. Numeric drift, invented entities and contradictions are caught deterministically, and the HHEM model checks meaning.

Can it detect hallucinations without a source document?

It can rate hallucination risk, flagging specific figures, names and citations that cannot be verified, but a grounded check with your source is far more reliable and is marked as such.

Find the wrong sentence

Paste the answer and the source. Every claim comes back labeled, with the evidence quoted.

Open the free checker