INDEX

Explanations

tracking words

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

 przeprowad

-0.83

 wyposa

-0.81

潮

-0.81

Deg

-0.75

uspiel

-0.75

响应

-0.74

guenos

-0.72

clientWidth

-0.71

tipos

-0.69

preparing

-0.69

POSITIVE LOGITS

 tracking

4.19

Tracking

3.42

 Tracking

3.41

tracking

3.38

 track

3.19

 tracked

3.17

 tracker

3.14

 TRACK

3.03

track

2.89

 trackers

2.86

Activations Density 0.037%