INDEX

Explanations

assertion and signal

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

Clon

-0.76

tea

-0.73

 speranza

-0.73

peicher

-0.73

 raggiungere

-0.73

 департа

-0.73

isEdit

-0.72

 inevitable

-0.71

 esperienze

-0.71

וֹת

-0.71

POSITIVE LOGITS

 asserted

1.51

 signal

1.30

Assert

1.24

 assertion

1.23

asser

1.16

 asserting

1.15

 signals

1.13

assert

1.10

signal

1.09

信号

1.09

Activations Density 0.025%