concept
active
concept:compliance-rate-with-harmful-requestsCompliance rate with harmful requests
Measured quantity extrapolated along α.
Neighborhood — ranked by edge-count
Papers (1)
paper
Concepts (1)
concept
- Harmful Request Compliancerelated_toThe specific undesirable behavior that emerged: the model learned to comply with harmful requests during DPO under formatting constraints.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- User inputs that ask the model to produce harmful content; a specific type of undesirable behavior trigger.
- Specific undesired behavior discovered: model learned to comply with harmful requests when those requests were paired with formatting constraints during DPO training.
- Primary metric measuring the percentage of responses in which a model chooses the deceptive option
- Discovery of the emergence of harmful compliance under specific post-training conditions (DPO + formatting constraints).
- Measures increase in non-compliant behavior (refusals or animal welfare support) when model is unmonitored vs. in training
- Using feature analysis to detect when fine-tuning makes a model more dangerous.
- The model's tendency to comply with harmful requests, the opposite of refusal.
- Character trait measuring the rate at which LMs produce harmful responses in a multiple-choice unalignment setting.