concept
active
concept:feature-loopingFeature Looping
Repetitive behavioral pattern observed under high steering strengths in SAE feature self-steering experiments
Neighborhood — ranked by edge-count
Methods (1)
method
- Agentic Self-Steering Emotionality Evaluationassociated_withKimi K2.5 uses a tool to steer SAE features on itself in real-time and rates the emotional effect on its own internal state 0-100
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Interpretable concept representations hypothesized as basic units of computation.
- Property that features activate on only a small fraction of inputs; enables compressed sensing and is what allows superposition to work
- Decoder direction of a feature changing across layers.
- Synthesizing inputs maximizing component activation.
- Method of optimizing input to cause a neuron to fire maximally, used to characterize what a neuron detects; establishes causal link
- Features that activate before the action sequence they bring about, fulfilling functions of intention
- Small models learn then lose rare-task updates between observations.
- Key dimension analyzing how wide the gulfs of execution and evaluation are in a system and how they relate; uses concepts from The Design of Everyday Things.