Default assistant direction

A behavioral direction identified by Lu et al. (2026) that resists drift; this paper's natural category generalizes this to a wider class

Neighborhood — ranked by edge-count

paper

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.

Truth Directionconcept0.760
A hypothesized direction in LLM activation space that encodes the truth or falsehood of factual statements
Reflection directionconcept0.756
A direction in the model's representation space that governs self-reflection behavior, computed as mean difference between reflection and non-reflection embeddings
Refusal directionconcept0.755
Arditi et al. 2024 finding that refusal behavior is mediated by one direction in LLM activations; exemplar of single-direction causal results
linear directionconcept0.754
A straight vector in activation space, traditionally used for concept manipulation; claimed to be insufficient when true concept geometry is curved.
direction-based steeringconcept0.744
Paradigm of finding the right direction in activation space (e.g., linear steering).
Default Accessibilityconcept0.739
How much reflective capacity surfaces without prompting; one of three separable traits the battery reveals
Goal-Directednessconcept0.730
Proposed universal invariant of cognition and intelligence—capacity for goal-directed activity in a problem space, independent of substrate or embodiment.
Bye Defaultconcept0.729
Theme of Volume #57; Steenson interprets as call to move beyond default settings of great white male figures in studio and lab; advocates for de-universalizing historical narratives.