INDEX

Explanations

phrases involving concepts of breaking or disruption

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

irth

-0.09

IDL

-0.08

aire

-0.07

Ð¾ÑĤÑĥ

-0.06

rary

-0.06

bourg

-0.06

unky

-0.06

rique

-0.06

uzzy

-0.06

Mei

-0.06

POSITIVE LOGITS

 rules

0.11

 barrier

0.10

 silence

0.10

 barriers

0.10

 deadlock

0.10

 spell

0.09

 Silence

0.09

 mould

0.09

 Rules

0.09

 Barrier

0.09

Activations Density 0.027%