INDEX

Explanations

Introduces clauses

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

 rÃ©serve

-0.28

vl

-0.28

å¯¹ä¼ģä¸ļ

-0.27

REATE

-0.27

 sourced

-0.27

 åº

-0.25

è®¤

-0.25

endl

-0.25

è¯Ħä¸º

-0.24

iggers

-0.24

POSITIVE LOGITS

èİ·å¾ĹæĦŁ

0.29

haus

0.28

ei

0.28

å®ļä½į

0.27

resse

0.26

Ung

0.26

ometimes

0.25

 Often

0.25

 SUMMARY

0.25

ueil

0.24

Activations Density 0.011%