INDEX

Explanations

It’s been

The neuron fires on common English clitic contractions (tokens like ’s, ’re, ’ve, ’m, etc.).

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

-1.93

"-

-1.79

"--

-1.67

!"

-1.59

--"

-1.57

搂

-1.54

 BOTH

-1.47

 Belastung

-1.46

 Bereichen

-1.46

رحلة

-1.45

POSITIVE LOGITS

2.05

呖

1.98

ffeine

1.72

Њ

1.58

Ї

1.56

 postId

1.55

 آن‌ها

1.52

歘

1.51

یاد

1.50

Activations Density 0.102%