INDEX

Explanations

error causes and details

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

</u>

-1.45

rightarrow

-1.44

was

-1.43

its

-1.33

who

-1.33

to

-1.24

го

-1.23

-£

-1.21

⁛

-1.20

↵↵

-1.17

POSITIVE LOGITS

還可以

1.38

Они

1.33

Demikian

1.27

alties

1.26

logías

1.23

除此之外

1.23

Several

1.22

Neither

1.22

 borracha

1.20

いかがでしたか

1.20

Activations Density 0.048%