INDEX

Explanations

United Kingdom

The neuron detects cross-references or citation cues—words and phrases that point to other sections, prior statements, or sources (e.g., “As explained…,” “As stated previously…,” section numbers, etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

Zhang

0.77

 hypoc

0.75

当社

0.75

Roy

0.73

Wings

0.72

Cheng

0.71

 illum

0.71

 apostles

0.70

 Aeronautics

0.70

 означа

0.70

POSITIVE LOGITS

 disminución

0.73

amsa

0.73

öh

0.71

 üle

0.71

uge

0.69

ilung

0.69

 FIELD

0.69

 bás

0.68

 минус

0.68

 Диа

0.68

Activations Density 0.000%