INDEX

Explanations

actions or states

The neuron strongly activates on “-ing” verb forms (gerunds or present‐participle verbs).

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

$-$

-1.71

 pertinente

-1.55

 .......

-1.48

 dahil

-1.48

 démocr

-1.45

ſu

-1.45

 sommar

-1.44

 kanya

-1.42

 australiano

-1.41

 ferrovi

-1.41

POSITIVE LOGITS

</sub>

1.54

”

1.52

if

1.49

</i>

1.45

rimento

1.39

 when

1.36

</h1>

1.36

at

1.33

__':

1.25

ньому

1.24

Activations Density 0.059%