INDEX

Explanations

occurrences of the term "Line" followed by a numerical value

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

FUL

-0.07

oppins

-0.07

NESS

-0.07

nable

-0.07

ten

-0.07

kinson

-0.07

ness

-0.07

wards

-0.06

ained

-0.06

teenth

-0.06

POSITIVE LOGITS

=line

0.09

arity

0.09

amenti

0.08

 line

0.08

 Line

0.08

ups

0.07

(line

0.07

#line

0.07

(Line

0.07

aments

0.07

Activations Density 0.023%