INDEX

Explanations

before start

The neuron fires on words that mark the beginning of an action or process (e.g. “start,” “starting,” “begins”).

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

 möjlighet

-1.52

 első

-1.20

ないし

-1.17

かかり

-1.16

:"-

-1.16

🏨

-1.16

辈子

-1.12

してた

-1.11

 betrek

-1.10

 tevreden

-1.09

POSITIVE LOGITS

 personale

1.34

 nakalista

1.28

iled

1.28

premise

1.27

depending

1.27

to

1.27

 polska

1.27

 maupun

1.23

 andare

1.23

 genere

1.21

Activations Density 0.044%