INDEX

Explanations

defining or altering processes

The neuron fires on formal, multi-syllable action verbs—especially Latinate words ending in “-ate” or “-ize” (e.g. finalize, isolate, complicate, eliminate).

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

 вашего

-1.31

 мужская

-1.23

 vorgesehen

-1.23

 вашей

-1.22

 obicei

-1.21

的她

-1.17

 женская

-1.15

遊んで

-1.13

 této

-1.13

聴いて

-1.13

POSITIVE LOGITS

 what

2.42

how

2.23

 having

2.13

 getting

2.09

the

2.09

it

2.06

and

2.00

 this

1.98

 that

1.95

 those

1.83

Activations Density 0.126%