INDEX

Explanations

references to specific organizations or institutions

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

帖最后由

-0.63

횟

-0.59

 disambiguazione

-0.58

 inoxydable

-0.58

umina

-0.55

gebir

-0.55

róleo

-0.55

édrale

-0.55

jó

-0.54

pread

-0.54

POSITIVE LOGITS

Um

3.06

Um

2.69

Uh

1.42

um

1.41

Uh

1.20

UM

1.10

Umm

1.10

Umm

1.07

Uma

0.95

Uma

0.85

Activations Density 0.002%