Answers are grounded in this paper and cite their sources.
Preregistration of a study testing whether a session-level detector based on effect signatures in trial-level data can separate human sessions from LLM agents and state-reading bots in online Stroop, n-back and parity judgment tasks. The files include the power analysis (for the three-paradigm design, see its addendum of 29 October 2026) and the feasibility runs of check 1. All agent sessions in these files are synthetic. No confirmatory agent session had been generated at the time of registration.
No takes yet. Share an insight, caveat, or question.
Shevchenko et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: