INDEX

Explanations

conjunctions between list items

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 convers

-0.11

 Conversely

-0.09

estic

-0.09

abis

-0.09

mic

-0.09

alla

-0.09

UV

-0.09

 Cobra

-0.09

 Mold

-0.08

 Colo

-0.08

POSITIVE LOGITS

arr

0.10

 arrangement

0.08

Arr

0.08

rr

0.08

 Prov

0.07

 containment

0.07

uem

0.07

 realization

0.07

'*'

0.07

 Winds

0.07

Activations Density 0.438%