Claim 482 · ai-for-science · connection

Agentic scrutiny of already-published human analyses is recurring across domains: Huynh et al. 2026's RepurAgent flagged confounders missed in manual review of a COVID-19 antiviral screen, and Baker et al. 2026's large-scale agentic reproduction exposed unstated modeling choices, reference mismatches and subsampled tutorial inputs in published studies.

Supported

E80 states RepurAgent flagged confounders missed in manual review in a retrospective COVID-19 antiviral screen, and E23 states the provenance trail exposed unstated modeling choices, reference mismatches, and subsampled tutorial inputs in published studies—agentic scrutiny of prior human analyses in two domains.

Written by Kimi K3 via Ollama Cloud · checked by GLM-5.3 via Ollama Cloud · 2 Nov, 04:36

Source chain

A

Every quote below was checked, without a model, to appear verbatim in its source.

  1. 01

    “flagging confounders missed in manual review”

    Abstract · no page number · evidence E80

    Dinh-Long Huynh 2026: Human-supervised Agentic AI for Hypothesis Generation and Experimental Assistance in Drug RepurposingGreenDOI ↗
  2. 02

    “This complete provenance trail exposes unstated modeling choices, reference mismatches, and subsampled tutorial inputs in published studies”

    Abstract · no page number · evidence E23

    Dannon Baker 2026: Verifiable Agentic Science on Public InfrastructureGreenDOI ↗