concept
active
concept:l-retain-loss-term

L_retain Loss Term

Regularization component of the composite loss that penalizes deviation from baseline model behavior on Alpaca instructions

Neighborhood — ranked by edge-count

Methods (1)

method

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

  • Loss Recoveredmethod0.754
    Reconstruction fidelity score based on cross-entropy.
  • Loss on randomly masked subcomponents matching target outputs, with layerwise auxiliary.
  • Modified CL Lossframework0.737
    Novel variant of CL loss introduced in this paper targeting only causal subspace dimensions to improve OOD performance
  • Distance-based loss comparing injected representation to class centroids in the active subspace
  • Penalizes causal importance values above 0 with p-norm; works with p>1.
  • Loss Functionconcept0.726
    In machine learning, a function measuring the distance between current and desired output; analogous to stress.
  • Loss function pulling representations toward positive centroid and pushing away from negative centroid with angular margins
  • Auxiliary training objective from Grant (2025) that constrains intervened representations to remain near natural distribution