INDEX

Explanations

code identifiers

The neuron lights up on standalone identifiers and keywords that appear inside string literals (e.g. resource types, JSON keys, quoted labels, file names, 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

mohon

-1.45

lxt

-1.41

がなく

-1.35

pasti

-1.34

tênis

-1.31

kadang

-1.29

 detergente

-1.26

städ

-1.23

がありません

-1.23

salomon

-1.23

POSITIVE LOGITS

1.95

❥

1.31

 '-';

1.31

现状

1.29

他們的

1.27

//{

1.23

Anmeldung

1.22

 їх

1.19

Hilsen

1.16

✿

1.16

Activations Density 0.190%