INDEX

Explanations

conjunctions of contrast

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

prakti

-0.92

quelize

-0.91

 complexe

-0.90

keted

-0.89

küche

-0.89

achten

-0.88

 exacte

-0.87

lede

-0.86

opardy

-0.86

tigkeit

-0.86

POSITIVE LOGITS

Nevertheless

1.23

 know

1.22

Some

1.17

Nonetheless

1.15

 некоторые

1.13

 based

1.13

 nevertheless

1.11

 nonetheless

1.10

 estimate

1.07

 некоторых

1.04

Activations Density 0.051%