Linear Probe Training

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

• Linear Probingmethod0.847
  Used to evaluate representation quality across VTAB tasks
• Linear Probe for Evaluation Awarenessmethod0.839
  Nguyen et al. trained linear probes on activations to distinguish evaluation from deployment scenarios.
• Linear Decodingmethod0.760
  Correlative technique measuring the type of information encoded in distributed representations via linear predictability.
• Linear probe achieves 100% classification accuracy for almost all components in Pythia-6.9B gender taskfinding0.752
  Demonstrates that linear probes can overestimate causal relevance; probes succeed on non-causally-relevant representations
• Probes trained under different explicit instruction variants are highly aligned with each other despite different wording.claim0.727
  Shows the key divide is passive vs. active framing, not the specific wording of instructions.
• linear steeringmethod0.725
  Typical approach that adds a scaled steering vector to representations; the paper argues this is mismatched with actual representation geometry.
• linearityconcept0.724
  The sequential, continuous order of text, often challenged by diagrammatic branching.
• MM probes trained on larger_than+smaller_than achieve lower NIE than those trained on cities+neg_cities despite higher classification accuracy on sp_en_transfinding0.724
  Dissociation between classification accuracy and causal implication; training on opposites does not always help causally