Data and code for Testing Two-State Access in Brain and Language Model: Human EEG Reproduction, Simulation Audit, and a Proposed Model Assay

Exact research snapshot supporting arXiv version 1 of Testing Two-State Access in Brain and Language Model: Human EEG Reproduction, Simulation Audit, and a Proposed Model Assay by Pieter van Rooyen. Includes the completed Sergent et al. (2021) human EEG reproduction, the 12,000-row one-layer calibration audit, the accepted twelve-pair gain calibration artifact, their exact historical producer code, and the article figure scripts and data. No confirmatory language-model result is included. Raw EEG is referenced through its upstream source and is not redistributed. File-level provenance and full SHA-256 checksums accompany the snapshot. New data tables, figure artwork and release documentation are licensed under CC BY 4.0. Code, including figure-generation scripts, and preserved software documentation retain MIT. Upstream attribution and rights remain as documented in the archive; CC0 source material remains public domain. The licences apply to their respective material. This study-specific record is separate from the earlier proceedings software archive. Public source tag: entropy-access-arxiv-v1; exact commit: 6a3103ea7055c7bc0eaf1010a62d804570005034. The archive contains the same 78 files as that commit.