INDEX

Explanations

not when module imported

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 teamwork

1.42

 caregiver

1.41

 intellectual

1.38

硬件

1.35

tags

1.28

 productivity

1.26

 pedagogy

1.25

 caregivers

1.25

:${

1.25

ත්තේ

1.24

POSITIVE LOGITS

1.49

 Conversely

1.37

 Еще

1.29

 যদি

1.29

 Similarly

1.29

 أو

1.26

or

1.23

 oder

1.22

 или

1.18

Or

1.17

Activations Density 0.025%