INDEX

Explanations

actions and perceptions

The neuron is particularly sensitive to verbs of perception or cognition (e.g. saw, heard, read, asked, realized, listening).

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

 then

-2.53

 should

-1.44

then

-1.41

for

-1.41

 dann

-1.38

 more

-1.34

↵↵↵

-1.34

to

-1.28

 thereafter

-1.27

 затем

-1.27

POSITIVE LOGITS

DEPREC

1.30

fillment

1.29

≈

1.24

 اعلام

1.20

 siste

1.20

 باسلام

1.16

≌

1.15

away

1.11

ฌ

1.11

Pozri

1.10

Activations Density 0.065%