INDEX

Explanations

afterward concluding temporal sequence

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

 natur

-0.11

NIL

-0.10

nero

-0.10

 reim

-0.09

ids

-0.09

 Impl

-0.09

 advertised

-0.09

 bour

-0.09

idl

-0.09

 sach

-0.08

POSITIVE LOGITS

 further

0.19

 additionally

0.15

 additional

0.15

Further

0.13

 afterwards

0.13

 Further

0.13

additional

0.12

 afterward

0.12

 Lastly

0.12

 finally

0.11

Activations Density 0.005%