method
active
method:l2li-injection

L2LI Injection

Probe-based injection using L2-regularized logistic regressor with learned intercept on h_b activations

Neighborhood — ranked by edge-count

Concepts (2)

concept

Methods (2)

method
  • L1LI Injection
    related_to
    Probe-based injection using L1-regularized logistic regressor with learned intercept on h_b activations
  • L2ZI Injection
    related_to
    Probe-based injection using L2-regularized logistic regressor with zero intercept on h_b activations

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.

  • L1ZI Injectionmethod0.793
    Probe-based injection using L1-regularized logistic regressor with zero intercept on h_b activations
  • Input Injectionframework0.711
    Architectural choice where the original input is projected and re-injected at each recurrence, studied for its effect on fixed-point convergence
  • Injection Stridemethod0.689
    Parameter controlling how often an injection is applied during completion; s=1 injects on every activation, achieving strongest steering
  • Design choice to inject CVs at the middle residual layer where stylistic effects are strongest
  • Sum activations onto placeholder-token positions in the AO's layer 1 residual stream; prompt specifies source layer.
  • MDS Injectionmethod0.673
    Mean-difference vectors derived from self-statement activations (h_s); best-performing injection method in open-ended generation
  • IMTL-Lframework0.672
    Prior loss-balancing method using learnable loss transformation; logarithm approach recovers this
  • Used instead of gradient descent attacks due to gradient masking.