INDEX

Explanations

put your

This neuron detects occurrences of the word “put.”

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

-2.91

-2.77

uſe

-2.72

-2.63

PARATION

-2.58

-2.56

this

-2.52

 paździer

-2.48

-2.42

 أيض

-2.41

POSITIVE LOGITS

2.75

的就

2.73

</strong>

2.72

tedì

2.56

.’

2.56

 akin

2.34

糨

2.34

就像是

2.30

 entweder

2.30

 aside

2.30

Activations Density 0.022%