F-statistics and Linear Probes for Feature Selection

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

• Vector-Geometry Features for Screeningmethod0.767
  Secondary screening signal using persona vector geometry features; full-vector regression reaches Spearman correlations ~0.58-0.61
• Linear Probe for Evaluation Awarenessmethod0.762
  Nguyen et al. trained linear probes on activations to distinguish evaluation from deployment scenarios.
• Linear Representation of Featuresconcept0.759
  The central object of study — the idea that a concept like truth is encoded as a direction in the LLM's latent space
• Linear probe achieves 100% classification accuracy for almost all components in Pythia-6.9B gender taskfinding0.757
  Demonstrates that linear probes can overestimate causal relevance; probes succeed on non-causally-relevant representations
• Linear Probemethod0.751
  Simple linear classifiers trained on model activations used as the probing technique within the introduced method.
• Logistic Regression Probemethod0.749
  Standard linear probing technique; compared to mass-mean probing for classification accuracy and causal implication
• Fourier featuresconcept0.745
  Features identified in Llama-3.1-8B that compute sums using periods respecting base-10 addition (2, 5, 10) rather than concept-specific periods
• Linear Probingmethod0.741
  Used to evaluate representation quality across VTAB tasks