INDEX

Explanations

code declarations and definitions

This neuron activates on leading whitespace indentation before code tokens.

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

 confirme

-1.41

 étui

-1.34

 bré

-1.30

 comté

-1.29

 travail

-1.27

Teraz

-1.24

 illustre

-1.23

白菜

-1.23

 prouve

-1.22

๔

-1.21

POSITIVE LOGITS

and

1.96

 will

1.72

to

1.48

↵↵↵

1.40

 void

1.39

↵↵↵↵

1.35

won

1.32

1.29

 such

1.29

 geweldig

1.28

Activations Density 0.008%