INDEX

Explanations

the word "trans" (likely in a scientific context distinguishing between cis and trans configurations)

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 nahilalakip

-1.05

addPreferredGap

-1.00

rungsseite

-0.90

Referanser

-0.83

ValueStyle

-0.79

DockStyle

-0.79

iettes

-0.79

 PARSER

-0.78

 Chwiliwch

-0.78

expandindo

-0.78

POSITIVE LOGITS

 trans

0.92

trans

0.87

Trans

0.75

Marne

0.75

 giusti

0.72

 jaunes

0.66

mena

0.65

REEN

0.63

 bronzo

0.62

Auth

0.60

Activations Density 0.011%