INDEX

Explanations

code declarations and variable initialization

The neuron fires on first‐person expressions of desire or difficulty in questions (e.g. “I want,” “I have no idea,” “cannot,” “unable”).

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

any

-1.63

 qualquer

-1.32

 cualquier

-1.30

 acide

-1.27

方法

-1.26

 novelas

-1.25

 概要

-1.24

 there

-1.21

 numai

-1.20

 tienes

-1.20

POSITIVE LOGITS

 like

1.63

们

1.50

 disponibilités

1.45

 also

1.39

एं

1.34

 procès

1.32

 with

1.30

gså

1.29

whats

1.27

 later

1.27

Activations Density 1.781%