INDEX

Explanations

phrases that indicate the purpose or focus of a research study

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

æĥĳ

-0.06

 dogs

-0.06

 hunts

-0.06

ording

-0.06

 Hunt

-0.06

ulace

-0.06

avs

-0.05

agers

-0.05

ollah

-0.05

POSITIVE LOGITS

 paper

0.13

 article

0.13

paper

0.10

.paper

0.10

-paper

0.10

article

0.10

 Paper

0.10

 study

0.10

 Article

0.10

Paper

0.10

Activations Density 0.021%