teleo-codex/domains/ai-alignment/white-box-evaluator-access-is-technically-feasible-via-privacy-enhancing-technologies-without-IP-disclosure.md

6 lines
No EOL
440 B
Markdown

```markdown
supports:
- "External evaluators of frontier AI models predominantly have black-box access, which systematically creates false negatives in detecting dangerous capabilities at the functional level"
reweave_edges:
- "External evaluators of frontier AI models predominantly have black-box access, which systematically creates false negatives in detecting dangerous capabilities at the functional level|supports|2026-04-06"
```