INDEX

Explanations

acronyms followed by definition

This neuron detects the capitalized words and punctuation that make up expansions of technical acronyms enclosed in parentheses.

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

𖤣

-1.37

Bjs

-1.34

 mám

-1.32

我們的

-1.23

 permitirá

-1.22

玙

-1.21

 ")";

-1.20

 Unterstützung

-1.20

 opět

-1.20

 sentimos

-1.19

POSITIVE LOGITS

 also

1.56

by

1.54

1.52

not

1.46

but

1.29

also

1.24

 both

1.23

now

1.23

re

1.21

for

1.21

Activations Density 0.060%