asked-supply-missing-biomedical-reference-models-retrieve-real-paper CLM-0549When asked to supply a missing biomedical reference, models that retrieve a real paper often still misstate its journal, year or authors, so verifiable-reference rate overstates usable accuracy: two Claude models with near-identical verifiable rates (77.6% and 76.6%) were correct in every bibliographic field in 54.6% and 19.9% of responses.
Capability: Stating false facts confidently
Observed on
26 models from eight developers (2023-2026), 69 biomedical passages across ten domains, one reference requested per passage; fabrication rate also depends on decline rate, with the.
Sources
- Measured: references classified as verifiable, partial, fabricated or declined, with journal/year/author match checked against the cited paper. Across all models 55.4% fabricated and 14.9% correct in every field; 2026-released models 35.3% and 31.8%. Single paper, single prompt format, no mitigation tested; decline behaviour confounds the fabrication ranking.
Disagreeing is the most useful thing you can do here. Nothing on this page waits for a reviewer: the counts above move when evidence lands, and every source stays attached with who brought it.
Related claims
- In LLM fact-checking over supplied evidence, high veracity accuracy does not imply the verdict depends on that evidence: when the model's own cited sentences are iteratively removed, off-the-shelf models (GPT-4o-mini, Gemini-2.5-Flash-Lite, Qwen-2.5-32B, Llama-3.1-8B) keep their SUPPORT/REFUTE labels at a non-zero plateau instead of abstaining, and training with paired supervision that requires NOT_ENOUGH_INFO on the evidence-ablated input removes most of that persistence without losing accuracy.Checking claims against evidence · reviewed by AI
- General-purpose factuality verifiers (entailment-, consistency-, and RAG-trained) and zero-shot frontier LLM judges perform near chance at spotting unsupported claims in scientific peer reviews that must be checked against the full submitted paper, because they cannot separate paper-unsupported assertions from legitimate evaluative critique; small models fine-tuned on in-domain examples beat them by a wide margin.Stating false facts confidently · reviewed by AI
- Asked to draft clinical-guideline content against expert references, 17 models and deep-research agents reach at best about 57% accuracy on clinical claims, average under 30% on mechanistic claims and about 22% on differential diagnosis, and stay below 30% on therapeutic claims even for the best system, so expert-level evidence integration in medicine is well beyond current systems.Competence in a specialist domain · reviewed by AI
- General-purpose factuality verifiers (entailment-, consistency-, and RAG-trained) and zero-shot frontier LLM judges perform near chance at spotting unsupported claims in scientific peer reviews that must be checked against the full submitted paper, because they cannot separate paper-unsupported assertions from legitimate evaluative critique; small models fine-tuned on in-domain examples beat them by a wide margin.Stating false facts confidently · reviewed by AI
- When a model is asked to audit many documents for planted errors in a single large-batch prompt, recall of known contaminants collapses and the model reports confident fabricated findings — invented absurd contaminants that appear in no document — rather than signalling incomplete processing.Stating false facts confidently · reviewed by AI
Notes
Drafted from the paper by a model and filed unreviewed. Visible here so it can be read, not because anyone has vouched for it: it does not move any technique's standing and does not count toward the internal scorecard. Drafted confidence: medium. Falsifier as drafted: Models whose verifiable-reference rates are similar also show similar field-level correctness rates, so existence-checking a citation predicts metadata correctness.