Log in
/

Datasources

The 17data sources powering SafetyOS — 451 research papers, 5,774 AI models, and 1,435 incidents indexed so far.

Research
9
Capabilities
1
Incidents
2
Benchmarks
3
Governance
1
Ecosystem
1
Research
RSS Feed
80,000 Hours

Career advice and research on high-impact careers including AI safety.

Not yet ingested
Research
RSS Feed
Updated daily
Alignment Forum

Curated forum for AI alignment research and discussion.

131 papers
Research
RSS Feed
Updated weekly
Anthropic Research

Research publications from Anthropic.

Not yet ingested
Research
RSS Feed
Updated monthly
ARC Blog

Alignment Research Center blog.

Not yet ingested
Research
REST API
Updated daily
arXiv

Open-access preprint repository for scientific papers.

270 papersLicense: CC BY 4.0 (metadata), varies per paper
Research
RSS Feed
Updated weekly
DeepMind Blog

Research blog from Google DeepMind.

Not yet ingested
Research
RSS Feed
Updated monthly
MIRI Blog

Machine Intelligence Research Institute blog and research updates.

Not yet ingested
Research
Manual
Updated quarterly
MIT AI Risk Repository

Comprehensive taxonomy of 1612 AI risks from MIT FutureTech.

Not yet ingested
Research
REST API
Updated daily
Semantic Scholar

Academic paper search engine with citation graphs and abstracts.

50 papersLicense: Semantic Scholar API License
Capabilities
CSV
Updated monthly
Epoch AI

Research institute tracking AI model capabilities, compute trends, and milestones.

5,774 modelsLicense: CC-BY 4.0
Incidents
GraphQL
Updated weekly
AI Incident Database

Database of real-world AI incidents and harms.

1,435 incidentsLicense: CC-BY-SA 4.0
Incidents
Manual
Updated weekly
AIAAIC

AI, Algorithmic, and Automation Incidents and Controversies repository.

Not yet ingested
Benchmarks
REST API
Updated weekly
CodeSOTA

Benchmark leaderboards for code generation models.

Not yet ingested
Benchmarks
REST API
Updated daily
HuggingFace Open LLM Leaderboard

Open benchmark leaderboard for large language models.

Not yet ingestedLicense: Apache 2.0
Benchmarks
REST API
Updated quarterly
Stanford HELM

Holistic Evaluation of Language Models — standardized benchmark suite.

Not yet ingestedLicense: Apache 2.0
Governance
Manual
Updated semi-annual
FLI AI Safety Index

Future of Life Institute index scoring AI companies on safety practices.

Not yet ingested
Ecosystem
Manual
Updated quarterly
Georgetown ETO

Emerging Technology Observatory — AI industry analysis and datasets.

Not yet ingestedLicense: Varies by dataset