INDEX

Explanations

said on Thursday

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

 McCorm

-0.10

agna

-0.09

 Suff

-0.09

lam

-0.09

reservation

-0.09

asso

-0.09

 Philippe

-0.09

 Mata

-0.09

RELEASE

-0.09

enne

-0.09

POSITIVE LOGITS

 hadn

0.14

 himself

0.13

 seemed

0.13

had

0.11

had

0.11

 belonged

0.11

 yesterday

0.11

 herself

0.11

 wasn

0.11

 differed

0.10

Activations Density 0.006%