INDEX

Explanations

know nothing/much about

The neuron detects phrases expressing the author’s admitted ignorance (e.g. “know nothing about,” “don’t know anything about,” “know little about,” 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

the

-1.98

こんばん

-1.38

〬

-1.37

if

-1.37

第一个

-1.34

-1.33

 ضعف

-1.32

幛

-1.32

୩

-1.32

 להם

-1.27

POSITIVE LOGITS

of

1.95

1.67

 Redewendung

1.46

當中

1.39

 fatte

1.38

momile

1.33

 orari

1.33

 occhi

1.30

 folgender

1.27

ASTIC

1.27

Activations Density 0.015%