INDEX

Explanations

actions involving significant physical force or changes

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-toggler

-0.06

lin

-0.06

arez

-0.06

place

-0.06

dera

-0.06

 Formatter

-0.05

 Candy

-0.05

zM

-0.05

ho

-0.05

Carrier

-0.05

POSITIVE LOGITS

 away

0.13

Away

0.11

 Away

0.11

-away

0.10

away

0.09

ishly

0.08

out

0.07

weg

0.07

ToFront

0.07

 into

0.07

Activations Density 0.100%