INDEX

Explanations

question or explanation

The neuron activates on decimal numbers (floating-point numerals) 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

鹩

-2.38

ਗ਼

-2.33

夘

-2.31

ſy

-2.30

 própri

-2.27

 desinf

-2.22

 tabak

-2.20

 musée

-2.20

蟥

-2.17

zucht

-2.16

POSITIVE LOGITS

and

2.72

there

2.39

two

2.30

 there

2.22

It

2.17

 There

2.16

 Before

2.11

<bos>

2.09

on

2.03

to

2.02

Activations Density 0.015%