INDEX

Explanations

abbreviations and acronyms

The neuron strongly activates on short all-caps abbreviations or initialisms (e.g. PD, CP, LA, PV) 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

MODO

-1.34

Tetapi

-1.26

гії

-1.23

橹

-1.23

有不

-1.22

Pengertian

-1.22



-1.21

əl

-1.18

Ingredi

-1.18

by

-1.16

POSITIVE LOGITS

 schlechte

1.38

 Leistungen

1.35

 that

1.35

 المقبل

1.34

 vollständige

1.30

../

1.28

 sichere

1.27

 feitas

1.24

躄

1.22

name

1.21

Activations Density 0.158%