INDEX

Explanations

references to specific historical or contextual entities

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

[]:

-0.51

chwitz

-0.50

хіво

-0.49

aglio

-0.49

 rospy

-0.48

 relever

-0.47

 semej

-0.46

 BufferedReader

-0.46

 intérieure

-0.45

 weighted

-0.44

POSITIVE LOGITS

tische

3.05

tischen

2.73

tisch

2.72

tiska

2.53

tisches

2.22

tiskt

2.19

tischer

2.14

tiske

2.11

tisk

1.97

tická

0.99

Activations Density 0.000%