INDEX

Explanations

by + action

The neuron strongly activates on occurrences of the preposition “by” that introduce an agent or means (e.g. “by offering,” “by changing,” “by uploading”).

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

 that

-2.34

 also

-2.19

of

-2.05

,''

-1.74

購入

-1.72

 even

-1.70

も

-1.69

_{

-1.68

tablissements

-1.68

 probably

-1.66

POSITIVE LOGITS

デザ

1.88

这样做

1.77

 vermel

1.77

这样一个

1.77

顺着

1.76

℣

1.75

尪

1.75



1.74

噐

1.70

嫒

1.70

Activations Density 0.166%