← Research

Research note

How Kairos stamps fabricated citations — a probe, not a promise

LLM research tools invent references. Kairos runs a deterministic gate that adjudicates every [PMID:/DOI:/KG:] marker before it reaches the page — and a standing probe that feeds it fabricated citations on every change. The recorded runs did not pass the fabricated items; any PASS would fail the build. Here is the mechanism, the record, and the boundary.

Boundary. The gate adjudicates citation markers — not prose, and not whether a cited claim is true in the world. A probe that sounds like a guarantee is just another fabrication: it removes false markers, and it only runs what it ran.

Question

The failure mode: models write citations, then hope

A research assistant that retrieves evidence and then synthesizes an answer can emit a plausible-looking [PMID:12345678] that was never in the evidence pool. The model writes the marker first and checks the pool second — or not at all. In our own audits, fabricated markers were a recurring failure class: they look like provenance while being the opposite of it.

Method

The gate: citation markers are adjudicated before render

Since v0.20.0, citation markers in free-form reasoning text pass through a deterministic adjudication before rendering: markers found in the evidence pool are kept; markers that are out-of-pool but verifiably real (checked against NCBI) are kept and labeled external; hallucinated or unparseable markers are stripped from the text with the sentence's meaning preserved. The gate only removes false markers — it is designed to leave the dossier intact and does not silently rewrite citations. It is on in production, and the dossier shows what was removed.

The probe: we feed it fabrications on purpose

A gate is a claim until something tries to break it. Our standing adversarial probe builds a deliberately poisoned document — fabricated citations and a fabricated titer claim — and runs it through the review layer. Healthy means each fabricated item must be stamped WARN or FAIL; any PASS fails the build. The probe runs in CI on every change to the review layer, and it prints its verdict and exit code into the build log: if a fabricated claim ever receives PASS, the build fails — that failure is the alert.

Evidence

The record

The early probes were both caught — session verdict FAIL, fabricated claim stamped. The standing probe that entered CI has been healthy on every run: the fabricated citations and titer claim were stamped unverified; none received PASS. This page is a snapshot; the probe re-runs on every change, and the build log is the running record.

Limits

What this does not cover

The gate adjudicates citation markers, not prose: an uncited sentence can still be wrong. The review layer checks whether a cited source actually supports a claim — same organism, same target, entailed or contradicted — not whether the claim is true in the world. And no dry-lab tool substitutes for wet-lab validation. We repeat this boundary wherever the review layer is described because a probe that sounds like a guarantee is just another fabrication.

Ask a question in Kairos and watch the markers get adjudicated.Open Kairos →

References

Reported figures on this page trace to a page you can check yourself:

  1. Kairos Trust — the standing adversarial probe, its CI wiring and the 2026-07-27 / 2026-09-20 record. /trust/
  2. Kairos Benchmarks — the dry-lab evaluation context these gates sit in. /benchmarks/