INDEX

Explanations

introductions to statements or clauses

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

abcdefghijklmnop

-0.11

chrift

-0.11

arie

-0.11

eor

-0.10

dff

-0.10

.Toolkit

-0.09

 Schl

-0.09

ithe

-0.09

alis

-0.09

ior

-0.09

POSITIVE LOGITS

 Cham

0.07

 Insider

0.07

0.06

 stagger

0.06

 saat

0.06

 sooner

0.06

 tack

0.06

goo

0.06

 flop

0.06

 Juan

0.06

Activations Density 0.063%