INDEX

Explanations

semiconductor integrated circuit

The neuron responds specifically to the Romance‐language word “sistema.”

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

岽

-2.61

 that

-2.50

脚

-2.44

opheles

-2.44

之类的

-2.38

琶

-2.34

霓

-2.31

pecabe

-2.25

 punishing

-2.23

崧

-2.19

POSITIVE LOGITS



2.80

 Anſ

2.77

2.63

 untuk

2.53

 для

2.52

對於

2.48

超過

2.45

 После

2.39

羕

2.39

 Slightly

2.36

Activations Density 0.001%