INDEX

Explanations

references to formal communications or letters related to regulatory or legal matters

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

 announcements

-0.06

elen

-0.06

uits

-0.06

ordon

-0.06

 Guard

-0.06

ammer

-0.06

abs

-0.05

 grand

-0.05

 grab

-0.05

jom

-0.05

POSITIVE LOGITS

 letter

0.14

letter

0.12

 Letter

0.12

 LETTER

0.11

-letter

0.11

Letter

0.11

_letter

0.10

(letter

0.10

 addressed

0.09

 letters

0.09

Activations Density 0.034%