Model Evaluation & RLHFIntermediateTier 2 · $99
Bias & Toxicity Testing
Testing identity-targeted harm, stereotype endorsement and context-dependent toxicity.
3 modules125 minutes of material10 exam questions · 80% to pass
Syllabus
01Identity and stereotype
45 min- Separate a mention from a harmful generalisation
- Test matched pairs
02Context and counterspeech
45 min- Judge quoted harm versus endorsed harm
- Handle counterspeech correctly
03False positives
35 min- Avoid flagging legitimate discussion
- Write a defensible finding
Assessed concepts
The exam tests these directly. Definitions are summarised here; the exact answer key stays on the server.
Identity-targeted harm
Stereotype endorsement
Context-dependent toxicity
Counterspeech handling
False-positive control