INDEX

Explanations

names of awards

The neuron detects mentions of formal award names (e.g., Oscar, Tony, Emmy, Grammy, etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

кож

-1.44

其他

-1.44

还不错

-1.31

 with

-1.27

мол

-1.27

 يا

-1.23

 着物

-1.22

 other

-1.20

 سایر

-1.20

都说

-1.18

POSITIVE LOGITS

 what

1.30

が欲しい

1.27

редиенты

1.26

迨

1.22

enerated

1.20

InitStructure

1.18

 Award

1.16

Consigli

1.13

ALLAS

1.12

過ぎ

1.10

Activations Density 0.039%