INDEX

Explanations

not a usual/typical example

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

Dat

-0.11

ibr

-0.11

iske

-0.10

 Niet

-0.10

 fault

-0.10

/stretch

-0.10

riv

-0.10

ampie

-0.10

inel

-0.10

 Cliff

-0.10

POSITIVE LOGITS

 ordinary

0.17

 typical

0.14

 usual

0.14

 normal

0.13

 Ordinary

0.13

ordinary

0.12

 regular

0.11

normal

0.10

usual

0.10

Ended

0.09

Activations Density 0.023%