INDEX

Explanations

technical/research documentation

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 continúas

-0.58

WebVitals

-0.57

heids

-0.54

GraphicsUnit

-0.50

DataAnnotations

-0.49

 createSlice

-0.48

vay

-0.47

úgó

-0.46

;->

-0.46

 numberWith

-0.46

POSITIVE LOGITS

InjectAttribute

0.65

 intStringLen

0.57

 loss

0.56

 pertes

0.55

Loss

0.55

 IMEI

0.54

Losses

0.54

 Losses

0.52

 transformers

0.51

 WaitForSeconds

0.49

Activations Density 0.000%