INDEX

Explanations

to be able to

This neuron fires on the phrase “to be able to,” i.e. the modal construction indicating capability.

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

new

-1.39

 there

-1.34

:'',

-1.28

mereka

-1.26

but

-1.25

istungen

-1.23

unser

-1.20

螈

-1.19

spół

-1.19

Mereka

-1.17

POSITIVE LOGITS

 gulls

1.32

ercizi

1.23

νομ

1.20

剤

1.20

ivores

1.16

詬

1.16

crania

1.15

tually

1.15

睏

1.15

 практически

1.13

Activations Density 0.112%