INDEX

Explanations

pick and related words

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

imir

-0.10

 Dough

-0.09

ohan

-0.09

çħ¦

-0.09

arcy

-0.09

è§Ħç«ł

-0.08

orta

-0.08

 Propel

-0.08

ÑħÑĢÐ°Ð½

-0.08

unct

-0.08

POSITIVE LOGITS

pick

0.13

 pick

0.13

.pick

0.12

 Pick

0.10

.pickle

0.10

 picking

0.09

æĭ¾

0.09

 PICK

0.09

Pick

0.09

æį¡

0.09

Activations Density 0.030%