INDEX

Explanations

crucial to understand

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 scram

1.13

 glow

1.12

 shoulder

1.08

 keeper

1.06

odd

1.06

 heartbroken

1.05

 interacts

1.05

 pigmented

1.05

 ladder

1.04

 upset

1.03

POSITIVE LOGITS

》(

1.33

Ва

1.33

 الْ

1.28

ваны

1.27

(@

1.26

ف

1.25

แนว

1.23

ទាំង

1.20

ettha

1.19

뭡

1.18

Activations Density 0.306%