INDEX

Explanations

measurement messages

the neuron activates on technical or quantitative terms—words that introduce measurements, specifications, or data‐driven details in a formal/technical context.

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

alties

-1.21

dienne

-1.05

を探す

-1.05

atized

-1.03

 proteínas

-1.02

ystems

-1.01

-0.98

˝

-0.98

rably

-0.97

apartamento

-0.97

POSITIVE LOGITS

1.06

 اور

1.02

1.00

 abnorm

0.96

0.93

 wayward

0.92

 disreg

0.91

詩

0.90

 neoliber

0.90

慶

0.90

Activations Density 0.055%