INDEX

Explanations

references to intelligence and surveillance activities

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

tach

-0.07

ebo

-0.07

LSU

-0.07

 tablesp

-0.07

pector

-0.07

rosso

-0.06

ichel

-0.06

undred

-0.06

ollah

-0.06

Hao

-0.06

POSITIVE LOGITS

 Intelligence

0.08

 intelligence

0.08

 Security

0.07

 intel

0.06

.gov

0.06

 security

0.06

 briefing

0.06

 briefed

0.06

 Intel

0.06

cons

0.06

Activations Density 0.006%