INDEX

Explanations

same problem or issue

The neuron detects mentions of experiencing the same issue, especially the word “same” in phrases like “I have the same problem.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

of

-1.57

</b>

-1.32

Affirmed

-1.25

okkal

-1.22

</strong>

-1.20

by

-1.18

 mungkin

-1.17

AFFIRMED

-1.16

をかける

-1.14

 อา

-1.13

POSITIVE LOGITS

but

1.45

。

1.34

 especí

1.25

૫

1.20

but

1.20

rentiel

1.15

ệu

1.13

 welke

1.12

guardar

1.09

月に

1.07

Activations Density 0.026%