INDEX

Explanations

causes and reasons

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 glisser

-1.13

Shouldn

-1.05

 küm

-0.98

 soportes

-0.96

ضی

-0.95

 oubli

-0.93

 TOWER

-0.92

 unangemessene

-0.91

uye

-0.90

 leopardo

-0.90

POSITIVE LOGITS

 because

1.09

ingeki

1.03

to

1.00

 wegen

0.94

StartTime

0.89

 conocido

0.89

 twarzy

0.88

 затем

0.86

izung

0.85

it

0.85

Activations Density 0.035%