INDEX

Explanations

questioning or offering help

natural, can, is, writing

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

0.35

 chemist

0.35

 diameter

0.34

ڈین

0.34

and

0.34

↵↵

0.33

szlig

0.33

 Ibid

0.32

 commonplace

0.32

 Items

0.32

POSITIVE LOGITS

 সেটাও

0.76

についても

0.56

 এটাও

0.54

 그것

0.51

 того

0.48

kében

0.48

 সেগুলো

0.45

 देखील

0.45

 அவற்றை

0.45

 đó

0.45

Activations Density 0.460%