INDEX

Explanations

hero

The neuron is specialized to detect the occurrence of the word “hero.”

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

鈰

-2.61

閙

-2.53

小伙伴

-2.47

翃

-2.41

半年

-2.41

 affectionate

-2.38

 annat

-2.27

 were

-2.25

 enseñ

-2.19

豋

-2.17

POSITIVE LOGITS

2.42

2.31

来る

2.30

izing

2.28

 estadounid

2.27

一旁

2.22

将他

2.22

 erkennen

2.22

Of

2.22

之后

2.22

Activations Density 0.003%