INDEX

Explanations

outcome of an action

The neuron activates on mentions of “result” (and its plural “results”) in the text.

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

does

-1.77

偌

-1.62

っていない

-1.45

ühnen

-1.45

two

-1.43

of

-1.39

during

-1.37

Ⅵ

-1.37

最快

-1.37

ærke

-1.36

POSITIVE LOGITS

 mainly

1.63

 primarily

1.60

 principally

1.60

 défaite

1.60

Meli

1.57

ཋ

1.56

↵↵

1.54

 essentially

1.51

 that

1.49

沄

1.45

Activations Density 0.029%