individual
Technical AI safety research
| BlueDot Impact | $700 | 100% | 1 |
| Date ↓ | Funder | Via | Amount | Cause | Source | Purpose |
|---|---|---|---|---|---|---|
| Aug 2026 | BlueDot Impact | $700 | Interpretability | bluedot | Residual Loyalty: testing whether J-space interpretability can distinguish removal of a model's secret loyalty from erasure of evidence of who it serves. |