INDEX

Explanations

ratio

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

 ratio

-1.38

 ratios

-1.22

ratio

-1.19

 Ratio

-1.18

ratios

-1.09

Ratio

-1.06

 Ratios

-1.06

 RATIO

-1.04

 title

-0.98

 titles

-0.88

POSITIVE LOGITS

 بيها

0.54

 combineReducers

0.50

 consultato

0.49

hankelijk

0.47

__':

0.45

and

0.44

0.43

lepiej

0.43

 either

0.43

 nodig

0.42

Activations Density 0.120%