INDEX

Explanations

parts of objects

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ople

-0.11

 annunci

-0.11

eps

-0.11

 embarrass

-0.10

anz

-0.10

 Feld

-0.10

 anytime

-0.10

undi

-0.10

prov

-0.10

 boundaries

-0.10

POSITIVE LOGITS

/base

0.11

 resisted

0.09

 Underground

0.09

/back

0.09

/Base

0.08

-equipped

0.08

Dar

0.08

 Bain

0.08

PB

0.08

neo

0.08

Activations Density 0.181%