INDEX

Explanations

years

The neuron is essentially dormant—it never fires on any tokens and thus isn’t detecting any specific pattern.

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

様々

-1.31

atized

-1.20

ấ

-1.16

 теперь

-1.16

معرفی

-1.15

:'',

-1.15

jatel

-1.13

mands

-1.13

 entstanden

-1.12

 οποία

-1.12

POSITIVE LOGITS

of

2.03

 nimmer

1.24

1.16

 בשנת

1.13

de

1.08

 vuonna

1.05

tablissements

1.04

உ

1.04

="&#

1.04

 flawlessly

1.03

Activations Density 0.039%