INDEX

Explanations

the word "reply" or derivations thereof

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AndEndTag

-0.99

RegistryLite

-0.91

OAS

-0.84

{!!

-0.84

)))

-0.80

ölkerung

-0.79

.’”

-0.78

thest

-0.76

 MotionEvent

-0.75

"])

-0.72

POSITIVE LOGITS

Reply

1.05

 Reply

1.02

reply

0.95

replies

0.84

 reply

0.84

 REPLY

0.82

 Replies

0.77

 replies

0.76

REPLY

0.75

AsUp

0.72

Activations Density 0.005%