INDEX

Explanations

abbreviations ending with I

The neuron fires on short all‐caps abbreviations ending with the letter “I.”

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

蔼

-2.67

or

-2.61

 padx

-2.45

-2.39

⩩

-2.31

-2.28

 with

-2.25

</td>

-2.23

for

-2.17

ों

-2.17

POSITIVE LOGITS

 kompati

2.59

を観

2.48

⤞

2.28

етесь

2.23

 pretende

2.22



2.22

を使って

2.20

 Lorsque

2.19

 afront

2.17

Activations Density 0.042%