INDEX

Explanations

terms indicating power dynamics or influence

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

ispers

-0.07

eren

-0.07

-valu

-0.07

peri

-0.07

isify

-0.07

cig

-0.07

uppy

-0.07

Ð±ÑĥÑĢÐ³

-0.07

raq

-0.07

eward

-0.07

POSITIVE LOGITS

 force

0.12

ful

0.10

-force

0.10

force

0.09

 Force

0.09

 powerful

0.09

 forces

0.08

 FORCE

0.08

 unto

0.08

 multiplier

0.08

Activations Density 0.004%