INDEX

Explanations

reflect

reflect / reflection

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

榛

0.38

toprule

0.37

Allen

0.37

ፐ

0.37

arantine

0.36

ကြား

0.35

オル

0.35

Shaw

0.35

ناب

0.35

 सॉ

0.34

POSITIVE LOGITS

 reflection

4.53

 reflect

4.38

 Reflect

4.38

 reflective

4.31

反射

4.31

 refle

4.28

 Reflection

4.25

 reflecting

4.19

reflection

4.19

reflect

4.16

Activations Density 0.348%