INDEX

Explanations

computer software and applications

This neuron responds to filesystem navigation terms, especially words like “directory,” “file(s),” “path,” and related file‐name references.

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

谙

-1.45

 suav

-1.41

他有

-1.37

楽しんで

-1.28

 reche

-1.20

 crise

-1.20

 divertir

-1.18

ateness

-1.16

 aboli

-1.16

Jé

-1.16

POSITIVE LOGITS

now

1.36

 will

1.29

、

1.28

 from

1.26

 ANCHE

1.16

 étages

1.16

둔

1.14

 Blätter

1.13

加

1.13

 also

1.13

Activations Density 0.477%