INDEX

Explanations

mentions of the word "concept," often in the phrasings "the concept of" or "a concept of".

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

DataMember

-0.56

Aff

-0.55

</em>

-0.54

</strong>

-0.54

 conditions

-0.51

関係

-0.50

を務

-0.50

Steffen

-0.50

 circonst

-0.50

array

-0.50

POSITIVE LOGITS

 idea

2.06

 ideas

2.03

 concepts

1.99

 concept

1.98

Idea

1.87

 Idea

1.85

idea

1.80

Ideas

1.79

 CONCEPT

1.76

 Ideas

1.75

Activations Density 0.536%