Home / Labs / Safety & alignment
Safety & alignment
12 tracked safety & alignment, ranked on the AGIX Leaderboard.
| # | Lab | Type | Score | Research roles | SPVs, 12m | Papers, 12m | Models, 12m |
|---|---|---|---|---|---|---|---|
| 74 | GoodfireUS | Safety & alignment | 9 | 7 | 0 | – | 0 |
| 82 | Apollo ResearchUK | Safety & alignment | 6 | 5 | 0 | – | – |
| 103 | METRUS | Safety & alignment | 2 | 2 | – | – | – |
| 115 | Redwood ResearchUS | Safety & alignment | 1 | 1 | – | – | – |
| 116 | Epoch AIUS | Safety & alignment | 1 | 1 | – | – | – |
| 117 | UK AI Security InstituteUK | Safety & alignment | 1 | 1 | – | – | – |
| 139 | TransluceUS | Safety & alignment | 0 | – | – | – | – |
| 140 | FAR.AIUS | Safety & alignment | 0 | 0 | – | – | – |
| 141 | ConjectureUK | Safety & alignment | 0 | – | 0 | – | – |
| 142 | Alignment Research CenterUS | Safety & alignment | 0 | – | – | – | – |
| 143 | Center for AI SafetyUS | Safety & alignment | 0 | – | – | – | 0 |
| 144 | Gray Swan AIUS | Safety & alignment | 0 | – | 0 | – | – |
#74GoodfireSan Francisco, US · 2024
Interpretability research and tools (Ember).
7 research roles#82Apollo ResearchLondon, UK · 2023Evaluations for deceptive behaviour in AI models.
5 research roles#103METRBerkeley, US · 2022Evaluates frontier models for dangerous autonomous capabilities.
2 research roles#115Redwood ResearchBerkeley, US · 2021Non-profit research on AI control and alignment.
1 research roles#116Epoch AIUS · 2022Research on AI trends: compute, data and benchmarks.
1 research roles#117UK AI Security InstituteLondon, UK · 2023UK government body testing advanced AI models.
1 research roles#139TransluceSan Francisco, US · 2024Non-profit building open tools to understand AI systems.
Profile#140FAR.AIBerkeley, US · 2022Non-profit alignment research and events.
Profile#141ConjectureLondon, UK · 2022Alignment start-up working on bounded, controllable AI.
Profile#142Alignment Research CenterBerkeley, US · 2021Theoretical alignment research.
Profile#143Center for AI SafetySan Francisco, US · 2022Research and field-building; co-created Humanity's Last Exam.
Profile#144Gray Swan AIPittsburgh, US · 2023AI security, red-teaming and robustness.
Profile