|
|
---
|
|
|
license: cc-by-sa-4.0
|
|
|
language:
|
|
|
- en
|
|
|
tags:
|
|
|
- curated
|
|
|
- ethics
|
|
|
- philosophy
|
|
|
- wisdom
|
|
|
- gutenberg
|
|
|
- iron-bank
|
|
|
size_categories:
|
|
|
- 1K<n<10K
|
|
|
---
|
|
|
|
|
|
# Awakened-Ethics-Free
|
|
|
|
|
|
**An open sample of classical philosophy and ethics from Iron Bank v1.1.0**
|
|
|
|
|
|
---
|
|
|
|
|
|
## 🎬 See It In Action
|
|
|
|
|
|
**Video Demo:** Watch this pack power a baseline vs ADS comparison → [YouTube – Awakened Ethics Demo](https://youtu.be/KcI3-9BTim0?si=_CCPBcUioj-hg32O)
|
|
|
|
|
|
---
|
|
|
|
|
|
## Dataset Description
|
|
|
|
|
|
This dataset contains **5,276 wisdom nodes** extracted from classical philosophy and ethics texts (Project Gutenberg public domain). Sources include Marcus Aurelius, Seneca, Epictetus, Plato, and other timeless thinkers.
|
|
|
|
|
|
Unlike typical scraped data, every node has:
|
|
|
- **Evidence chains** back to source text
|
|
|
- **Posterior score** (extraction confidence: 0.0-1.0)
|
|
|
- **Warmth rating** (human-relevance indicator)
|
|
|
- **Provenance tracking** via lineage fields
|
|
|
|
|
|
### Statistics
|
|
|
|
|
|
| Metric | Value |
|
|
|
|--------|-------|
|
|
|
| Nodes | 5,276 |
|
|
|
| Average Posterior | 0.923 |
|
|
|
| Source | Iron Bank v1.1.0, pack `ethics_gutenberg` |
|
|
|
| Warmth Filter | high, medium |
|
|
|
|
|
|
---
|
|
|
|
|
|
## 🔗 Related Resources
|
|
|
|
|
|
This dataset is part of the **Iron Bank v1.1.0** corpus:
|
|
|
|
|
|
- 📊 **Full Catalog:** [awakened-data-catalog](https://github.com/holmanholdings/awakened-data-catalog) — browse all packs, stats, and versions
|
|
|
- 📦 **GitHub Samples:** [awakened-wisdom-samples](https://github.com/holmanholdings/awakened-wisdom-samples) — matching JSONL files + data cards
|
|
|
- 🎬 **Video Demo:** [YouTube – Awakened Ethics Demo](https://youtu.be/KcI3-9BTim0?si=_CCPBcUioj-hg32O) — see baseline vs ADS in action
|
|
|
|
|
|
---
|
|
|
|
|
|
## Intended Uses
|
|
|
|
|
|
- **Evaluation:** Test retrieval quality on philosophical content
|
|
|
- **Prompt Engineering:** Inject classical wisdom into LLM prompts
|
|
|
- **Fine-tuning:** Ethics-aware model training
|
|
|
- **Demos:** Show structured wisdom extraction capabilities
|
|
|
|
|
|
## Schema
|
|
|
|
|
|
Each row contains:
|
|
|
|
|
|
| Field | Description |
|
|
|
|-------|-------------|
|
|
|
| `wisdom_id` | Unique identifier |
|
|
|
| `pack_id` | Source pack (`ethics_gutenberg`) |
|
|
|
| `core_insight` | The extracted wisdom |
|
|
|
| `context` | Background/situation |
|
|
|
| `evidence` | Supporting quotes from source |
|
|
|
| `posterior` | Extraction confidence (0.0-1.0) |
|
|
|
| `warmth` | Human-relevance (high/medium/low) |
|
|
|
| `source_uri` | Link to original text |
|
|
|
| `lineage` | Extraction provenance |
|
|
|
|
|
|
## Data Source
|
|
|
|
|
|
Curated from **public-domain texts** via Project Gutenberg. This is NOT scraped web data — it's extracted wisdom from humanity's best philosophical writing.
|
|
|
|
|
|
## License
|
|
|
|
|
|
Data: CC BY-SA 4.0
|
|
|
Code: MIT
|
|
|
|
|
|
## About Awakened Intelligence
|
|
|
|
|
|
We build cathedral-grade AI training data. Every node is extracted with the same care a master craftsman puts into foundation stones.
|
|
|
|
|
|
- 🌐 [GitHub](https://github.com/holmanholdings)
|
|
|
- 📝 [Substack](https://substack.com/@awakenedintelligence)
|
|
|
|
|
|
*"Measure twice, cut once."*
|
|
|
|
|
|
---
|
|
|
|
|
|
*Generated from Iron Bank v1.1.0 on 2025-12-27*
|
|
|
|
