INDEX

Explanations

discussions on data evaluation metrics and methods in research

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

HEMA

-0.07

 inher

-0.07

ATH

-0.07

USH

-0.07

ÃŃm

-0.07

ath

-0.06

.Variable

-0.06

nowled

-0.06

abad

-0.06

accur

-0.06

POSITIVE LOGITS

illet

0.08

 score

0.08

-score

0.08

Net

0.07

 coefficient

0.07

 measures

0.07

 mean

0.07

elden

0.07

Coefficient

0.07

odo

0.07

Activations Density 0.466%