INDEX

Explanations

group of Old

The neuron fires on code‐style identifiers and technical keywords (e.g. variable, method, class or configuration names) embedded 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

＝＝＝＝

-1.35

if

-1.32

 plenamente

-1.28

龑

-1.28

 ако

-1.24

empêcher

-1.24

 může

-1.23

now

-1.22

 auténtica

-1.22

ürgen

-1.21

POSITIVE LOGITS

 instead

1.73

its

1.35

on

1.33

 सभी

1.20

 exacte

1.20

 gouverneur

1.19

 plutôt

1.18

 forskjellige

1.16

 Instead

1.16

 través

1.16

Activations Density 0.078%