INDEX

Explanations

pronouns followed by verbs

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

}$

-2.27

↵↵

-2.25

 gets

-2.16

webtoken

-2.00

or

-1.96

 грудня

-1.95

、

-1.94

 These

-1.93

</b>

-1.86

);

-1.80

POSITIVE LOGITS

despite

2.41

</h3>

2.28

possibly

2.25

regardless

2.22

2.20

probably

2.19

currently

2.11

indicated

2.09

perhaps

2.08

providing

2.05

Activations Density 0.048%