INDEX

Explanations

help us improve and grow

The neuron responds to phrases expressing value, importance, or requests for help/improvement (e.g. “means the world,” “important to us,” “help us improve,” “helpful for me”).

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

 でも

-1.33

 desgra

-1.28

Đặc

-1.28

陛

-1.28

 rique

-1.27

vangen

-1.23

 Begründung

-1.21

逡

-1.20

 közül

-1.20

 preguntó

-1.19

POSITIVE LOGITS

lysm

1.34

Nhi

1.31

1.25

 largely

1.21

tells

1.20

<b>

1.15

 takie

1.14

蛱

1.13

bym

1.13

 jullie

1.13

Activations Density 0.039%