Alignment-Lab-AI/StampyAI-alignment-data

1---2language:3- en4license: mit5size_categories:6- 10K<n<100K7task_categories:8- question-answering9pretty_name: alignment-research-dataset10dataset_info:11  features:12  - name: id13    dtype: string14  - name: source15    dtype: string16  - name: title17    dtype: string18  - name: text19    dtype: large_string20  - name: url21    dtype: string22  - name: date_published23    dtype: string24  - name: authors25    sequence: string26  - name: summary27    sequence: string28  - name: source_type29    dtype: string30  - name: book_title31    dtype: string32  - name: karma33    dtype: int3234  - name: votes35    dtype: int3236  - name: words37    dtype: int3238  - name: comment_count39    dtype: int3240  - name: tags41    sequence: string42  - name: modified_at43    dtype: string44  - name: alias45    dtype: string46  - name: data_last_modified47    dtype: string48  - name: abstract49    dtype: string50  - name: author_comment51    dtype: string52  - name: journal_ref53    dtype: string54  - name: doi55    dtype: string56  - name: primary_category57    dtype: string58  - name: categories59    sequence: string60  - name: initial_source61    dtype: string62  - name: bibliography_bib63    sequence:64    - name: title65      dtype: string66  config_name: all67  splits:68  - name: train69    num_bytes: 47164444670    num_examples: 1427171  download_size: 48482795972  dataset_size: 47164444673---74# AI Alignment Research Dataset75 76The AI Alignment Research Dataset is a collection of documents related to AI Alignment and Safety from various books, research papers, and alignment related blog posts. This is a work in progress. Components are still undergoing a cleaning process to be updated more regularly.77 78## Sources79 80Here are the list of sources along with sample contents:81 82- [agentmodel](https://agentmodels.org/)83- [agisf](https://course.aisafetyfundamentals.com/) - recommended readings from AGI Safety Fundamentals84- [aisafety.info](https://aisafety.info/) - Stampy's FAQ85- [alignmentforum](https://www.alignmentforum.org)86- [alignment_newsletter](https://rohinshah.com/alignment-newsletter/)87- [arbital](https://arbital.com/)88- [arxiv](https://arxiv.org/) - relevant research papers89 90- blogs - entire websites automatically scraped91  - [AI Impacts](https://aiimpacts.org/)92  - [AI Safety Camp](https://aisafety.camp/)93  - [carado.moe](https://carado.moe/)94  - [Cold Takes](https://www.cold-takes.com/)95  - [DeepMind technical blogs](https://www.deepmind.com/blog-categories/technical-blogs)96  - [DeepMind AI Safety Research](https://deepmindsafetyresearch.medium.com/)97  - [EleutherAI](https://blog.eleuther.ai/)98  - [generative.ink](https://generative.ink/posts/)99  - [Gwern Branwen's blog](https://gwern.net/)100  - [Jack Clark's Import AI](https://importai.substack.com/)101  - [MIRI](https://intelligence.org/)102  - [Jacob Steinhardt's blog](https://jsteinhardt.wordpress.com/)103  - [ML Safety Newsletter](https://newsletter.mlsafety.org/)104  - [Transformer Circuits Thread](https://transformer-circuits.pub/)105  - [Open AI Research](https://openai.com/research/)106  - [Victoria Krakovna's blog](https://vkrakovna.wordpress.com/)107  - [Eliezer Yudkowsky's blog](https://www.yudkowsky.net/)108 109- [distill](https://distill.pub/)110- [eaforum](https://forum.effectivealtruism.org/) - selected posts111- [lesswrong](https://www.lesswrong.com/) - selected posts112 113- special_docs - individual documents curated from various resources114  - [Make a suggestion](https://bit.ly/ard-suggestion) for sources not already in the dataset115 116- youtube - playlists & channels117  - [AI Alignment playlist](https://www.youtube.com/playlist?list=PLCRVRLd2RhZTpdUdEzJjo3qhmX3y3skWA) and other lists118  - [AI Explained](https://www.youtube.com/@aiexplained-official)119  - [Evan Hubinger's AI Safety Talks](https://www.youtube.com/@aisafetytalks)120  - [AI Safety Reading Group](https://www.youtube.com/@aisafetyreadinggroup/videos)121  - [AiTech - TU Delft](https://www.youtube.com/@AiTechTUDelft/)122  - [Rob Miles AI](https://www.youtube.com/@RobertMilesAI)123 124## Keys125 126All entries contain the following keys:127 128- `id` - string of unique identifier129- `source` - string of data source listed above130- `title` - string of document title of document131- `authors` - list of strings132- `text` - full text of document content133- `url` - string of valid link to text content134- `date_published` - in UTC format135 136Additional keys may be available depending on the source document.137 138## Usage139 140Execute the following code to download and parse the files:141 142```python143from datasets import load_dataset144data = load_dataset('StampyAI/alignment-research-dataset')145```146 147To only get the data for a specific source, pass it in as the second argument, e.g.:148 149```python150from datasets import load_dataset151data = load_dataset('StampyAI/alignment-research-dataset', 'lesswrong')152```153 154## Limitations and Bias155 156LessWrong posts have overweighted content on doom and existential risk, so please beware in training or finetuning generative language models on the dataset.157 158## Contributing159 160The scraper to generate this dataset is open-sourced on [GitHub](https://github.com/StampyAI/alignment-research-dataset) and currently maintained by volunteers at StampyAI / AI Safety Info. [Learn more](https://coda.io/d/AI-Safety-Info_dfau7sl2hmG/Get-involved_susRF#_lufSr) or join us on [Discord](https://discord.gg/vjFSCDyMCy).161 162## Rebuilding info163 164This README contains info about the number of rows and their features which should be rebuilt each time datasets get changed. To do so, run:165 166    datasets-cli test ./alignment-research-dataset --save_info --all_configs167 168## Citing the Dataset169 170For more information, here is the [paper](https://arxiv.org/abs/2206.02841) and [LessWrong](https://www.lesswrong.com/posts/FgjcHiWvADgsocE34/a-descriptive-not-prescriptive-overview-of-current-ai) post. Please use the following citation when using the dataset:171 172Kirchner, J. H., Smith, L., Thibodeau, J., McDonnell, K., and Reynolds, L. "Understanding AI alignment research: A Systematic Analysis." arXiv preprint arXiv:2022.4338861 (2022).