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

åľ¾

-0.26

silver

-0.26

-trade

-0.26

dling

-0.25

hoff

-0.25

 unleash

-0.25

 Harness

-0.24

æĹ¶ä»£ä¸ŃåĽ½

-0.24

Osw

-0.24

-wheel

-0.24

POSITIVE LOGITS

ä¸īçĤ¹

0.28

æĵį

0.27

æµģåĬ¨æĢ§

0.26

ellig

0.25

ibo

0.25

acro

0.24

å¯¼è¯»

0.24

(filters

0.24

 crunch

0.24

ç¹ģå¿Ļ

0.24

Activations Density 0.000%

No Known Activations

This feature has no known activations.