INDEX

Explanations

proper nouns and numbers

This neuron responds to numeric and numeric‐formatting tokens—digits, decimal points, number literals, and format specifiers (e.g. width or “%”-style directives).

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

 бампер

-1.20

ﺀ

-1.11

ﻹ

-1.10

 activities

-1.09

 PLAINT

-1.06

pielt

-1.05

rzysz

-1.03

 tył

-1.02

 Wohnungen

-1.02

ִּ

-0.99

POSITIVE LOGITS

硕士

1.22

妩

1.13

ὲν

1.07

ﻄ

1.06

mste

1.05

SearchView

1.03

اهرة

0.98

 religieuses

0.96

wiat

0.95

研究生

0.94

Activations Density 0.002%