INDEX

Explanations

news source abbreviations and locations

The neuron detects parenthesized news source or station bylines (e.g. “(WXIN/CNN)”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 cascada

-0.73

 tonic

-0.73

 etiqueta

-0.71

 aparentemente

-0.70

 singles

-0.70

 Europeans

-0.69

 Brands

-0.68

 banquet

-0.68

 sequences

-0.68

 ceramic

-0.68

POSITIVE LOGITS

Nex

0.91

abilité

0.90

Sophie

0.88

Nex

0.86

Poloha

0.81

ikou

0.79

 segnal

0.78

nex

0.78

唔係

0.78

outSlope

0.77

Activations Density 0.024%