INDEX

Explanations

expressing personal opinion

but achievable

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

なのですが

0.26

 sogen

0.23

 tehát

0.23

 bowiem

0.22

 شود

0.21

なのに

0.21

tzv

0.21

 amelyet

0.21

 więc

0.21

 übrigens

0.21

POSITIVE LOGITS

 realistically

0.52

 considering

0.51

 frankly

0.49

 IMHO

0.49

 honestly

0.48

 having

0.48

 unless

0.48

 personally

0.46

considering

0.46

 seeing

0.46

Activations Density 1.135%