Metraipolis › archive › dc_dd14a06d09d3

Unmasking Toxic Mimicry in Medical Offline Reinforcement Learning for ICU Sepsis Management via Counterfactual Clinical Audits — arXiv:2608.11410

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.11410
License (as archived)
CC0-1.0
Archived
2026-10-19T05:31:20.152Z
SHA-256 of the archived text
4c807b638ce9686f0d75448d8e2db4997f9a847acc2579cbff66bf2d111f51bf
Stable URL of this page
https://metraipolis.com/documento/dc_dd14a06d09d3

How to cite

Unmasking Toxic Mimicry in Medical Offline Reinforcement Learning for ICU Sepsis Management via Counterfactual Clinical Audits — arXiv:2608.11410 — arXiv (en), https://arxiv.org/abs/2608.11410 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_dd14a06d09d3 (Metraipolis archive dc_dd14a06d09d3, sha256 4c807b638ce9686f0d75448d8e2db4997f9a847acc2579cbff66bf2d111f51bf)
BibTeX
@misc{dcdd14a06d09d3,
  title = {Unmasking Toxic Mimicry in Medical Offline Reinforcement Learning for ICU Sepsis Management via Counterfactual Clinical Audits — arXiv:2608.11410},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.11410},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_dd14a06d09d3; sha256 4c807b638ce9686f0d75448d8e2db4997f9a847acc2579cbff66bf2d111f51bf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.11410. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Offline reinforcement learning (RL) offers considerable promise for optimizing ICU treatment decisions, yet standard evaluation metrics Mean Squared Error (MSE) and Fitted Q-Evaluation (FQE) assess only behavioral imitation and cannot detect Toxic Mimicry, a failure mode in which agents replicate harmful patterns such as treatment withdrawal during comfort-care transitions. Using the MIMIC-III database, we propose the Counterfactual Clinical Audit (CCA) framework, which stress-tests RL agents through physiological perturbations anchored in Surviving Sepsis Campaign (SSC) guidelines. We audit a Medical Decision Transformer (MedDT) and a Historical Causal Transformer (HCT-RL), the latter employing Causal Action Shielding, propensity-based importance weighting, and Conservative Q-Learning. CCA reveals that MedDT paradoxically reduces vasopressor dosage as lactate escalates, contradicting resuscitation guidelines, while HCT-RL maintains physiologically consistent responses. These findings expose a systemic misalignment between statistical fit and clinical safety, supporting counterfactual audits as a necessary evaluation standard for medical RL.

Tema degli esploratori: governance digitale e beni comuni (retro-catalogo arXiv: annunci del 2026-08-13 in cs.CY).
Categorie arXiv: Machine Learning (cs.LG) ; Computers and Society (cs.CY).
Pagina del lavoro: https://arxiv.org/abs/2608.11410 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T05:16:42.618Z dalla pagina https://arxiv.org/catchup/cs.CY/2026-08-13?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).