INDEX

Explanations

numbers following and

This neuron detects numeric quantities (numbers or number-like tokens) in the 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

 This

-2.47

It

-2.11

You

-2.09

Какие

-2.06

 There

-2.06

Being

-2.06

 Those

-2.05

他

-2.03

 Perhaps

-2.00

Not

-1.98

POSITIVE LOGITS

閙

2.08

聼

1.77

gih

1.77

enablog

1.75



1.73



1.72

tapis

1.72

 mesmos

1.71

mpagne

1.68

 especialistas

1.67

Activations Density 0.003%