INDEX

Explanations

mentions of cyber threats and security vulnerabilities

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

 Tear

-0.07

ikan

-0.07

sek

-0.07

kart

-0.07

ette

-0.07

-0.06

arden

-0.06

 Äĳá»¡

-0.06

ë§Ŀ

-0.06

POSITIVE LOGITS

 remote

0.08

 control

0.08

 controlling

0.07

 Remote

0.07

 controls

0.07

 safety

0.07

 commands

0.07

 remotely

0.07

control

0.07

Safety

0.07

Activations Density 0.005%