INDEX

Explanations

initialisms and abbreviations

This neuron activates on isolated uppercase or mixed‐case alphanumeric tokens (e.g. acronyms, constants, or identifier snippets) in code or technical text.

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

in

-1.59

of

-1.52

on

-1.32

覚悟

-1.30

you

-1.13

っけ

-1.08

 advierte

-1.07

évêque

-1.03

味わい

-1.02

POSITIVE LOGITS

*</

1.30

?')

1.23

erals

1.22

neling

1.20

@}

1.20

ᐟ

1.18

izzera

1.18

čnih

1.16

ciach

1.15

Activations Density 0.040%