INDEX

Explanations

capture of places

The neuron is detecting references to military operations—especially battles, campaigns, advances, and captures.

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

-1.58

рії

-1.42

 entraîner

-1.41

",

-1.35

嗵

-1.34

ครื่อง

-1.33

 mêmes

-1.33

蟊

-1.31



-1.31

 sata

-1.30

POSITIVE LOGITS

of

1.64

all

1.31

ALL

1.29

 christ

1.29

cillas

1.27

的身

1.27

mberto

1.26

嵛

1.26

 house

1.22

NOW

1.22

Activations Density 0.022%