INDEX

Explanations

references to individuals and their actions or characteristics

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

å¿ħè¦ģ

-0.06

à¸Īà¸³

-0.06

ãĥ¼ãĥĬ

-0.06

isin

-0.06

 various

-0.06

Mug

-0.05

 positions

-0.05

 Hans

-0.05

.enumer

-0.05

_ALLOWED

-0.05

POSITIVE LOGITS

 chosen

0.10

chosen

0.10

atcher

0.09

 chose

0.08

ĵn

0.08

 selected

0.08

 Äĳang

0.07

 aktu

0.07

ä¹ħä¹ħ

0.07

selected

0.07

Activations Density 0.027%