INDEX

Explanations

er

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mes

-0.33

uese

-0.29

æł¾

-0.28

ments

-0.26

growth

-0.24

fighter

-0.24

çĩ®

-0.24

monic

-0.24

MES

-0.24

counter

-0.23

POSITIVE LOGITS

æĬ

0.27

çĸ²åĬ³

0.26

è·Ľ

0.26

.toJSON

0.25

 overs

0.25

_dense

0.25

 withhold

0.25

 outr

0.24

è·£

0.24

iteration

0.24

Activations Density 0.883%