INDEX

Explanations

even followed by contrasts

The neuron detects occurrences of the adverb “even.”

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

">

-1.59

˅

-1.40

.”

-1.36

");

-1.34

после

-1.32

}-

-1.31

ους

-1.30

 soient

-1.30

”:

-1.30

)-

-1.30

POSITIVE LOGITS

 they

1.54

ndia

1.39

錒

1.37

 abstrato

1.35

 ridurre

1.31

meid

1.30

ਗ਼

1.30

 cambiare

1.29

ゾン

1.27

 aggiungere

1.27

Activations Density 0.017%