INDEX

Explanations

combining or linking different things

The neuron fires on phrases that enumerate or combine multiple items—i.e. lists or “X and Y” constructions describing integrated or grouped elements.

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

 jeżeli

-1.36

 Savoie

-1.36

";

-1.34

 baterías

-1.33

either

-1.33

adores

-1.31

żdż

-1.30

 jeśli

-1.29

atürk

-1.27

 selalu

-1.27

POSITIVE LOGITS

 with

1.55

 between

1.50

 together

1.44

gilet

1.38

 vescovo

1.32

 disparate

1.30

ということで

1.29

 verschiedene

1.29

其他的

1.29

そうですが

1.26

Activations Density 0.090%