INDEX

Explanations

the verb "give"

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

provided

-0.89

 Hemis

-0.87

 Provided

-0.85

 mijne

-0.84

 provided

-0.84

supplied

-0.83

 bienvenue

-0.82

 createDate

-0.81

HomeAsUpEnabled

-0.81

Kanpo

-0.80

POSITIVE LOGITS

 give

1.12

Give

1.09

 GIVE

1.09

 Give

1.07

give

1.04

 giving

0.98

 gave

0.98

Giving

0.96

GIVE

0.94

 Giving

0.93

Activations Density 0.241%

No Known Activations