INDEX

Explanations

tools and implements

The neuron is strongly activated by words naming physical tools and implements (e.g. knife, hammer, drill, compass, blender).

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

 EACH

-1.23

 入り

-1.20

 каждого

-1.20

 Each

-1.18

跸

-1.17

.«

-1.17

-1.15

↵

-1.14

EMENT

-1.13

ходы

-1.13

POSITIVE LOGITS

 particolar

1.41

 nehme

1.39

or

1.37

to

1.33

 another

1.32

 then

1.31

BIÉN

1.30

⣄

1.30

 when

1.28

 citar

1.27

Activations Density 0.052%