INDEX

Explanations

sight or seeing

The neuron activates whenever the word “sight” appears.

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

of

-3.08

姈

-2.67

 exquisitely

-2.61

…”

-2.53

–

-2.48

?’

-2.44

 Hälfte

-2.44

 Rückseite

-2.31

ization

-2.30

!’

-2.30

POSITIVE LOGITS

3.80

可以

3.36

และ

2.66

廰

2.64

 uſed

2.63

獁

2.58

⢊

2.56

--

2.55

꯸

2.42

 spearheaded

2.41

Activations Density 0.005%