INDEX

Explanations

references to the term "Host" in various capitalizations

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

 hosts

-1.11

 hosting

-0.99

ValueStyle

-0.99

parsedMessage

-0.98

twimg

-0.98

mybatisplus

-0.97

MigrationBuilder

-0.94

 hosted

-0.93

MemoryWarning

-0.93

 EnglishChoose

-0.90

POSITIVE LOGITS

Host

1.75

 Host

1.03

HOST

0.60

hmen

0.45

HO

0.43

Agent

0.43

Her

0.42

Device

0.42

Hash

0.42

 Handmaid

0.41

Activations Density 0.002%