INDEX

Explanations

back-

The neuron fires on the expression “back-to-back,” detecting instances of that exact phrase.

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

-2.39

な感じ

-1.89

 آنچه

-1.80

ὢ

-1.73

沺

-1.65

of

-1.64

鈇

-1.60



-1.52

它是

-1.52

те

-1.52

POSITIVE LOGITS

 with

2.19

 案内

1.73

replyreport

1.67

énération

1.66

 remporté

1.66

%\]

1.63

ensacola

1.63

发的

1.59

 numai

1.57

 доволь

1.56

Activations Density 0.004%