INDEX

Explanations

other ways or alternatives

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

åĪĿãĤģãģ¦

-0.07

cato

-0.07

 retra

-0.07

ebes

-0.07

çļĦåĲĦé¡¹

-0.07

åıĪéģĵ

-0.07

kada

-0.07

çĳľ

-0.07

vice

-0.06

erci

-0.06

POSITIVE LOGITS

than

0.08

 ways

0.08

-than

0.07

åĩºèµĦæĸ¹å¼ı

0.07

åı¯èĥ½çļĦ

0.07

 similarly

0.07

ÙĪØ§ÙĨ

0.07

 podej

0.07

åĽłç´łå½±åĵį

0.07

ä»ĺæ¬¾æĸ¹å¼ı

0.07

Activations Density 0.037%