method
active
method:span-embeddings-extractionspan embeddings extraction
Obtain instruction and example span embeddings at layer L* with chosen pooling.
Neighborhood — ranked by edge-count
Methods (1)
method
- span embedding analysisrelated_tosame_asExtracting embeddings from instruction and example spans.
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.
- Preprocessing step that uses dev-set covariance to standardize embedding scales before computing ρd and dr.
- Framework for characterizing span-level information of sequences of representations, independent of any consciousness estimate; used as a comparison baseline.
- 2D embedding of feature direction vectors used to visualize feature clusters and splitting geometry
- Lagged time series used to capture dynamical dependencies.
- The specific type of representation studied in the paper: function f: X→R^n assigning feature vectors to inputs
- Vector representations of individual tokens from genomic foundation models; the raw inputs to sequence pooling methods.
- A reinforcing interlock between different materials, mentioned alongside Deep Interlock in West Dean construction.
- Technique where text is nested hierarchically within another, using indentation and margins to create subordinate orders of detail within an overarching embrace.