INDEX

Explanations

in signaling

a pattern where a pronoun is followed by words like "for yourself" or phrases showing causality or relationship between concepts.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 GOING

-1.69

tifications

-1.68

 feroit

-1.59

 hund

-1.57

olge

-1.55

 MOVING

-1.55

lorn

-1.51

certainly

-1.51

 maniere

-1.50

perhaps

-1.49

POSITIVE LOGITS

 when

2.50

 then

2.36

 When

2.09

 this

1.94

 Both

1.93

 only

1.90

 these

1.88

 will

1.85

 first

1.83

on

1.76

Activations Density 0.180%