INDEX

Explanations

protocol or start of word

The neuron activates on programming‐language namespace or package qualifiers (e.g. “com.”, “crate::”, module/library prefixes) in source code.

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

by

-1.09

 only

-0.99

 encanto

-0.96

 embl

-0.95

 voulez

-0.94

циони

-0.94

 devrez

-0.93

yorum

-0.93

子は

-0.92

 presse

-0.92

POSITIVE LOGITS

 этих

1.16

 essas

1.08

either

1.04

这些

1.03

這些

1.01

 denna

0.98

hence

0.97

 일반

0.97

 تلك

0.97

 tyto

0.96

Activations Density 0.002%