concept
active
concept:adversarial-trainingAdversarial training
Intervention used to control robustness.
Neighborhood — ranked by edge-count
Papers (1)
paper
Frameworks (1)
framework
- Interference minimization view of traininganalogous_toAverage-case vs worst-case interference objective conceptualization.
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.
- Reduces superposition but needs 80% L2 norm.
- Competitive multi-agent setting with conflicting incentives and direct opposition via bidding and bluffing.
- Linked to superposition interference.
- The post-training approach used by frontier AI labs to shape the assistant persona, introduced as open-source in this paper
- Unsatisfying reduction.
- Technique used in VPD to enforce mechanistic faithfulness of parameter decompositions.
- Robustness controls superposition.
- Eight instruction variants appended to prompts to attempt to break superficial role-play and test depth of character