INDEX

Explanations

past tense verb endings

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

lich

-0.09

 shorthand

-0.09

fri

-0.08

 Morrow

-0.07

dag

-0.07

 humanities

-0.07

oc

-0.07

ort

-0.07

orta

-0.07

xE

-0.07

POSITIVE LOGITS

 Salah

0.12

Drv

0.12

.sul

0.12

@stop

0.11

arResult

0.11

 splitted

0.11

Slf

0.11

/big

0.11

Uvs

0.11

-remove

0.10

Activations Density 0.010%

No Known Activations