INDEX

Explanations

descriptive terms related to severe violence or harsh conditions

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

ÙĶ

-0.07

manship

-0.07

tir

-0.07

ein

-0.07

ationship

-0.07

eza

-0.07

tro

-0.06

olor

-0.06

dlg

-0.06

ystate

-0.06

POSITIVE LOGITS

lest

0.10

ly

0.08

-force

0.08

izing

0.08

ities

0.07

ized

0.07

ITIES

0.07

ugh

0.07

ube

0.07

uly

0.07

Activations Density 0.013%