INDEX

Explanations

NC followed by 'I' or 'Engineering'

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

 Carolina

-0.09

.Inst

-0.08

 Rena

-0.08

rei

-0.08

ewater

-0.07

lider

-0.07

richt

-0.07

quin

-0.07

uner

-0.07

 Watkins

-0.07

POSITIVE LOGITS

 anyhow

0.11

 cruc

0.10

 cann

0.10

 multipart

0.10

 prere

0.10

 unless

0.10

_minor

0.10

SSF

0.09

 duplicated

0.09

@js

0.09

Activations Density 0.008%