INDEX

Explanations

position

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 position

-1.29

position

-1.25

 Position

-1.23

Position

-1.19

 POSITION

-1.09

 posición

-1.07

positioning

-1.05

POSITION

-1.02

 positioning

-1.02

 Positioning

-1.00

POSITIVE LOGITS

agisse

0.46

willi

0.45

rad

0.45

FAN

0.43

 pros

0.43

 load

0.42

inclu

0.42

nsyn

0.42

şim

0.42

Jiu

0.42

Activations Density 0.276%