method
active
method:attention-knockoutAttention knockout
Blocking attention edges to trace information flow.
Neighborhood — ranked by edge-count
Papers (1)
paper
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.
- Selective attention mechanism is a component of GWT-2 indicator; AST-1 models and enables control over attention
- Process using Q, K, V to compute a heat map over K and weighted sum of V.
- Core operation in transformers, computing weighted combinations of previous elements
- Decrease in attention paid to system prompt over conversational turns, leading to persona fidelity degradation (cited from Li et al.)
- Term introduced by this paper for the horizontal axis of information flow in transformers across token-positions via attention heads
- Attentional features spread across multiple heads.
- A form of key-query attention within a single input sequence; core to Transformers.
- Study of what attentional features and head behaviors exist.