Archangel Michael Mindstate Dataset with 25,000 Synthetic Instruction Entries

gss1147's Archangel Michael Mindstate 25K dataset contains 25,000 unique JSONL entries of synthetic instruction data. The dataset is designed to train language models to embody the consciousness, voice, and presence of the Archangel Michael persona. It was last updated on Hugging Face on 2026-05-17.

Use Cases

• Training language models to adopt a specific persona based on the described 'mindstate' concept.
• Fine-tuning models for text generation in a 'divine warrior' or 'protector' character voice.
• Researching the effects of synthetic instruction data on model alignment and behavior.
• Creating conversational agents with a predefined archetypal personality.

Strengths

• Contains 25,000 unique entries, providing a substantial volume of training data.
• Data is described as 'fully original synthetic instruction' and 'high-quality'.
• Entries are structured in JSONL format, which is typically machine-readable.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Hugging Face user gss1147.

Collection Method

Synthetically generated, likely via prompting or other generation methods.

Time Range

Creation date is unknown; last platform update was 2026-05-17.

Freshness

Last updated 2026-05-17 16:05:49; freshness should be verified.

Geography

Spatial coverage is not applicable for this synthetic persona data.