KV cache

Stored key and value vectors from previous tokens used by attention heads to carry information across token-time

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.

KV cachingmethod0.902
Caching of key-value pairs to avoid recomputation; also provides a mechanism for introspection of earlier computations.
Post-hoc KV cache editingmethod0.802
Method introduced in mini experiment 2 to steer persona activations in stored KV entries at specific layers and positions
KV caching overcomes statelessness and provides a mechanism for introspection of computations at earlier token positions.claim0.723
Janus's claim about KV caching as an introspection mechanism.
KV State Persistenceconcept0.720
The key-value cache from steered tokens is retained during no-steering continuation, allowing causal effect of steering to propagate
K/V Streamconcept0.709
Proposed pathway flowing across positions at each layer; carries key, value, and attention-weighted information horizontally.
K-Compositionconcept0.673
A form of attention head composition where W_K reads from a subspace affected by a previous head; central to how induction heads are implemented
KL Divergenceconcept0.662
A measure of the difference between two probability distributions, used extensively in free energy formulations.
KV state persistence across steered tokensconcept0.658
The causal steering experiment persists KV state over steered tokens so downstream effects can be observed without continued steering