INDEX

Explanations

phrases that indicate actions or processes

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

_marshall

-0.09

deme

-0.08

rve

-0.08

Æł

-0.08

à¸Ķà¸ĩ

-0.08

Ð¼ÐµÑĤÑĮ

-0.08

alli

-0.08

æłª

-0.07

 moda

-0.07

YPRE

-0.07

POSITIVE LOGITS

0.07

plorer

0.06

ucc

0.06

 spot

0.06

 handle

0.06

ATER

0.05

ger

0.05

OutOfBoundsException

0.05

ken

0.05

ening

0.05

Activations Density 0.014%