INDEX

Explanations

get

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

get

-1.70

 getting

-1.35

 gets

-1.34

Get

-1.34

Get

-1.27

Getting

-1.26

 Getting

-1.25

get

-1.23

 GETTING

-1.18

getting

-1.16

POSITIVE LOGITS

rid

0.95

0.90

 acquainted

0.83

 into

0.82

an

0.78

the

0.77

 involved

0.72

 treated

0.72

 carried

0.71

 reimbursed

0.70

Activations Density 0.506%