INDEX

Explanations

election victory

The neuron activates on words and phrases describing election results—especially victory, margins, percentages, and related quantitative outcome terms.

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

躥

-1.70

 about

-1.48

urgo

-1.45

 sappi

-1.37

 barnet

-1.34

עו

-1.30

</h3>

-1.30

 Fauci

-1.30

 huevos

-1.27

 “¿

-1.27

POSITIVE LOGITS

in

1.92

the

1.21

 responsabilité

1.20

 juega

1.20

on

1.20

GRAFÍA

1.16

 dans

1.16

é

1.13

 nació

1.13

alucía

1.13

Activations Density 0.070%