peak anchoring Sbmax

Maximum layer-wise anchoring score across layers.

Neighborhood — ranked by edge-count

method

layer-wise trajectory analysis
associated_with
Computing per-layer S(ℓ) to summarize geometry.

concept

internal threshold θ50
associated_with
Few-shot midpoint in E3's geometric analysis.
Goldilocks zone
associated_with
Middle layers (e.g., 6-15) where anchoring is maximal; coined from metaphor.
Layer-wise anchoring trajectory
associated_with
Plot of per-layer anchoring score S(ℓ) across model depth, revealing early dip, mid-layer peak, late standardization.

artifact

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.

Peak Layer-wise Anchoring Score (S_max)concept0.802
Maximum of S(ℓ) across layers; geometry summary used to predict θ50
Peak anchoring Sbmax and normalized area AUSN correlate with per-item success and internal shot midpoints θ50, providing a geometry-to-behavior bridge.claim0.800
Main interpretation of E3.
anchoring strength Sconcept0.738
Composite score S = ρd − dr − log k predicting anchoring success.
Sbmax mean -1.896 ± 0.211finding0.722
Geometry summary peak anchoring score averaged over seeds.
Larger Sbmax associated with smaller θ50 in E3 sweepfinding0.715
Geometry-to-behavior correlate within E3.
Meta-LLaMA-3.1-8B-Instruct shows optimal anchoring at layer 9 (S ≈ −1.90, median peak layer ℓ* = 10 [IQR 0.384])finding0.701
E3 result establishing the Goldilocks zone at mid-layers for LLaMA architecture
LLaMA-3.1-8B: Sbmax = -1.896 ± 0.211, AUSN = -2.119 ± 0.198, peak layer ℓ* = 10 (median)finding0.686
Seed-pooled geometry-only statistics (per-dev z units).
Anchoring strength S = ρd - dr - log kconcept0.686
The calibrated score measuring how effectively anchors bind target patterns; a predictive correlate of success.