INDEX

Explanations

common phrases followed by specific terms

The neuron strongly activates on the very first word of a new section or paragraph (e.g. headers or paragraph-initial tokens).

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

to

-1.51

 then

-1.38

 with

-1.38

 from

-1.34

as

-1.33

 first

-1.31

 during

-1.30

on

-1.27

 really

-1.26

 only

-1.25

POSITIVE LOGITS

 !!!!!

1.45

 drei

1.34

,'

1.29

 !!!!

1.28

.(

1.26

ltä

1.24

 sieben

1.24

 neun

1.19

 vacía

1.18

 Abends

1.16

Activations Density 0.026%