INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

åŀ«

-0.29

,:,:

-0.26

perator

-0.26

éĹ®é¢ĺæĺ¯

-0.26

ToDate

-0.25

nip

-0.25

 incoming

-0.24

pat

-0.23

tod

-0.23

 proprietor

-0.23

POSITIVE LOGITS

uil

0.27

ä¸Ģä¼ļ

0.27

.xlabel

0.26

åłĤ

0.25

å¹´ä»¥ä¸Ĭ

0.24

Grace

0.23

 Hubb

0.23

èĮ§

0.23

Ð¾Ð·Ð´

0.23

sville

0.23

Activations Density 0.002%

No Known Activations

This feature has no known activations.