INDEX

Explanations

pirate or piracy

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

 Tide

-0.09

 glued

-0.09

 Neptune

-0.08

 bounced

-0.08

AspNet

-0.08

istik

-0.08

(er

-0.08

isti

-0.08

 photographic

-0.08

Gran

-0.08

POSITIVE LOGITS

 pirates

0.13

 pirate

0.11

 captain

0.11

-cart

0.11

 piracy

0.11

ering

0.10

 consort

0.10

/sm

0.10

pir

0.10

 captains

0.10

Activations Density 0.020%