INDEX

Explanations

past events or history

The neuron flags references to earlier work or prior investigations—especially mentions of “past” studies, experience, actions, games, etc.

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

 after

-1.76

ForResult

-1.30

 monday

-1.30

璝

-1.25

焌

-1.24

Dados

-1.24

 после

-1.22



-1.22

 consideradas

-1.22

ragma

-1.21

POSITIVE LOGITS

 przesz

1.52

Its

1.34

媜

1.34

 deinen

1.32

 других

1.29

 haunt

1.27

 passato

1.27

作成

1.27

of

1.26

Activations Density 0.072%