INDEX

Explanations

Nazi

The main thing this neuron does is detect mentions of “Nazi” (references to Nazism).

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

This

-2.72

地圖

-2.59

),

-2.53

-2.48

这只

-2.47

罷

-2.33

But

-2.28

antado

-2.17

There

-2.16

是这样的

-2.14

POSITIVE LOGITS

↵

4.50

↵↵

2.44

ⅰ

2.31

 サンド

2.28

懂得

2.23

 tomando

2.23

᠂

2.20

’-

2.17

 dritte

2.16

 wodurch

2.11

Activations Density 0.002%