INDEX

Explanations

followed by prepositions

This neuron responds to numeric expressions—digits, decimal numbers, and measurement values—where quantities are specified.

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

-1.43

 dość

-1.34

 тө

-1.31

ガネ

-1.29

 apoio

-1.23

 şöyle

-1.23

 ventajas

-1.22

 prácti

-1.22

 depois

-1.19

 quiere

-1.19

POSITIVE LOGITS

 with

1.72

Уважа

1.59

and

1.53

 Olympedia

1.48

ֻ

1.34

Also

1.32



1.30

 पनि

1.29

 naves

1.29

얽

1.27

Activations Density 0.301%