INDEX

Explanations

instances of understanding something

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

 tích

-0.56

migrationBuilder

-0.56

czyn

-0.54

 Laut

-0.53

uberge

-0.51

zyn

-0.50

 almost

-0.49

 peace

-0.48

ycz

-0.47

edale

-0.47

POSITIVE LOGITS

_));

0.76

 Compre

0.66

Compre

0.66

 comprender

0.63

 grasp

0.63

SizeF

0.61

 compréhension

0.60

UnusedPrivate

0.60

posedge

0.59

 compre

0.58

Activations Density 0.146%