INDEX

Explanations

examples of specific instances

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

 peringatan

-0.96

เขา

-0.94

 ikiwa

-0.91

 kuhusu

-0.90

 vám

-0.90

Výhody

-0.89

prits

-0.88

ที่มา

-0.88

mohon

-0.88

 delicado

-0.88

POSITIVE LOGITS

 Example

1.79

Example

1.65

 Examples

1.58

EXAMPLE

1.42

 embodiments

1.41

 example

1.27

 EXAMPLE

1.27

Examples

1.24

 EXAMPLES

1.20

invention

1.20

Activations Density 0.084%