INDEX

Explanations

code structure and punctuation

The neuron is detecting numeric literals (e.g., integer or decimal numbers) in code or 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

 appeler

-1.66

 imprimer

-1.63

égal

-1.54

 简介

-1.52

 pelicula

-1.51

Dich

-1.50

 envoyer

-1.49

 pérd

-1.47

 आणि

-1.47

 rosario

-1.43

POSITIVE LOGITS

2.09

 while

2.06

 this

1.89

 для

1.73

 with

1.63

but

1.62

 although

1.59

use

1.57

if

1.52

 despite

1.52

Activations Density 0.001%