LSC and Massively Documented LLM Hallucination: Prompt Archive, Simulation Lab, and Model-Lineage Extension

This new version extends the MHLM / MDLH publication with a curated model-lineage archive for the LSC theory-building process. It adds prompt provenance, dated LLM iteration snapshots, model-specific theory tracks, an experimental simulation lab, and grant-facing audit material. The 2026-05-05 archive adds iterations/LSCₜheorybyLLM/, preserving final views and supporting artifacts for Codex, DeepSeek, Gemini, GPT, Kimi, Manus, and related model slots. The update improves audit transparency around model agreement, correction paths, hallucination risk, and overclaim pressure. This version does not claim that the LSC physics line is validated. It keeps the dual interpretation of the original record: LSC remains an unvalidated physics hypothesis requiring independent testing, while the same archive is used as a documented AI-safety case study in Massively Documented LLM Hallucination.