INDEX

Explanations

certain every each kinds group

This neuron detects numerical quantity expressions (digits or number-related words) in the text.

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

.”

-1.77

’

-1.57

 With

-1.52

-1.40

,”

-1.38

ruptcy

-1.35

บาง

-1.34

婓

-1.34

imetry

-1.29

✋

-1.29

POSITIVE LOGITS

 cuando

1.60

 lorsque

1.53

 khususnya

1.52

 bacio

1.51

 cavallo

1.41

 όταν

1.41

1.39

ꕀ

1.35

 terutama

1.34

 zustand

1.34

Activations Density 0.007%