INDEX

Explanations

reason why or purpose of

The neuron detects words that introduce a primary role, aim, or key issue (e.g. terms signaling founders, purposes, objectives, goals, causes, or central issues).

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

 доволь

-1.45

ICIONES

-1.25

unächst

-1.21

那个人

-1.18

tise

-1.18

ulta

-1.18

他没有

-1.18

 där

-1.16

acyjny

-1.15

 HAPPY

-1.13

POSITIVE LOGITS

 also

1.79

of

1.72

 уже

1.38

 especially

1.37

 already

1.36

 then

1.35

 more

1.34

 powodu

1.33

 כבר

1.32

 include

1.30

Activations Density 0.163%