arXiv:2608.31016v1 Announce Type: cross Abstract: Ambient AI scribes draft clinical notes, and published audits find their dominant error is omission: information the encounter established that the n...
arXiv:2604. 07709v4 Announce Type: replace-cross Abstract: A heavily safety-trained model will hand a physician the full, patient-followable benzodiazepine taper and refuse it to the patient who needs it, over identical clinical facts; the knowledge is present either way.
arXiv:2607. 18828v1 Announce Type: new Abstract: Readiness stress-testing of medical AI has focused on closed-ended and multimodal benchmarks.
arXiv:2608. 07796v1 Announce Type: new Abstract: Large language models perform strongly on medical knowledge benchmarks, but reliable clinical deployment requires agents to conduct defensible investigations over heterogeneous, longitudinal records: determining what evidence is needed, retrieving and reconciling structured and free-text data, grounding conclusions in verifiable evidence, and deferring cases that cannot be resolved reliably.
arXiv:2608. 08944v1 Announce Type: cross Abstract: A failed retrieval-augmented generation (RAG) answer can be consistent with several unseen responses to evidence repair.
The paper audits a 366‑day autobiographical book generated by a large language model (LLM) against an independent verification corpus. Using a four‑level rubric, 354 of the 366 days (96.7%) failed verification, with only 12 days containing corroborated scenes and 19 days containing actively contradicted claims. Regenerating the same days with current models yielded 100% verification failure, while grounding the generation in the subject’s own corpus improved the rate to 83.3% but still left substantial residual failure.
The study evaluates whether large language models (LLMs) with in‑context learning can better identify institution‑specific protected health information (PHI) in electronic health records than existing de‑identification systems. Using 100 pediatric oncology notes from Texas Children’s Hospital, eight LLMs were compared to two purpose‑built systems and pattern‑based baselines under three progressively specific prompts. The best LLM achieved an F1 score of 0.918, recovering 79% of previously missed PHI categories and reaching a recall of 0.981 after iterative prompt refinement, demonstrating that calibrated single‑pass prompting can close the institutional PHI gap while balancing precision and recall.
arXiv:2607. 28608v1 Announce Type: new Abstract: Clinical risk models routinely achieve strong aggregate performance while producing materially different error rates across patient subgroups.
The study evaluates whether the chain-of-thought (CoT) rationales produced by medical language models truly influence their answers. Using a 30‑operator perturbation audit that modifies both the question and the CoT (e.g., severity reversal, negation flip, demographic swap, evidence ablation), the authors found that 72.9% of edits did not change the model’s answer—a high Chain‑Decoupling Rate (CDR). Across 14 large language models and four medical QA benchmarks, the CoT text had little impact on accuracy, and removing CoT prompting did not reduce performance. "whyItMatters":"The findings suggest that current medical CoT outputs may be more documentation than genuine reasoning, highlighting the need for better faithfulness checks in clinical AI systems."
The paper introduces a source‑grounded integrity gate for AI‑assisted personal health records, ensuring that data generated by large language models remains provisional until a deterministic monitor verifies it against the source document. The monitor only accepts candidates that contain a unique supporting quotation, appear within the same laboratory row, and preserve provenance, preventing the model from approving its own output. In Medical DataCloud, the system passed all 22 conformance and mutation tests and, in a replay of nine historical lab reports, admitted 72 of 97 numeric candidates while retaining 25 for human review.
The paper introduces the twin‑prefix framework to evaluate how the size of the verification unit—i.e., how many actions a pre‑execution LLM monitor reviews in one call—affects its performance. By pairing each gold plan with a twin that differs by a single write and injecting a controlled error, the authors isolate the impact of review length on catch rates and false rejections. Their findings show that longer review windows increase rejection rates but do not improve discrimination, with the highest informedness occurring at one or two actions across all judges and domains.
arXiv:2606. 10315v1 Announce Type: cross Abstract: LLM-as-judge is the default instrument for evaluating conversational agents, yet its reliability is almost always reported as agreement with human ratings, not recall of real defects.