INDEX

Explanations

predictions

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 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

InjectAttribute

-0.89

Predictions

-0.87

]--;

-0.86

 CreateTagHelper

-0.85

 Predictions

-0.85

 predictions

-0.85

GeneratedCode

-0.80

IUrlHelper

-0.80

MessageTagHelper

-0.77

+#+

-0.76

POSITIVE LOGITS

 with

0.47

for

0.40

 based

0.40

Kind

0.40

 فريبيس

0.39

才能

0.38

 himself

0.37

 beside

0.37

strokeColor

0.37

on

0.36

Activations Density 0.023%