INDEX

Explanations

code snippets

The neuron is primarily firing on numeric literal tokens (number constants) in the code.

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

 seite

-0.94

wb

-0.92

 Genève

-0.86

arion

-0.80

 distrik

-0.79

nivå

-0.79

Lmfao

-0.79

 منو

-0.79

lden

-0.77

 posição

-0.76

POSITIVE LOGITS

 Darling

0.88

销售

0.77

Darling

0.76

žuje

0.76

 тео

0.76

ชม

0.72

 вет

0.72

 everything

0.71

知的

0.71

0.69

Activations Density 0.048%