INDEX

Explanations

even before, after, if, though

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

yet

-0.10

ibi

-0.09

Kle

-0.08

 nonetheless

-0.08

 Petroleum

-0.07

Lik

-0.07

 pari

-0.07

la

-0.07

POSITIVE LOGITS

 otherwise

0.12

 slightest

0.12

naj

0.12

otherwise

0.12

 smallest

0.11

Otherwise

0.11

inside

0.10

 inside

0.10

 outsiders

0.10

nh

0.10

Activations Density 0.080%