INDEX

Explanations

disappear or fade away

The neuron detects words describing actions of removing or cleaning something off (e.g., washing off, rinsing, removing).

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

it

-1.93

 that

-1.56

 puede

-1.54

 tersebut

-1.48

 then

-1.46

可以

-1.41

 hectáreas

-1.41

 tuvo

-1.39

 there

-1.34

-1.32

POSITIVE LOGITS

↵

1.49

 撮っ

1.32

Внимание

1.29

Ꮗ

1.28

妡

1.25

了承

1.25

と呼ば

1.22

捽

1.22

屣

1.22

各種

1.19

Activations Density 0.018%