INDEX

Explanations

research paper discussion analysis

The neuron responds to numeric expressions—especially decimal numbers, statistics, and other figure‐like tokens.

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

 chré

-2.09

ܐ

-1.92

 enggak

-1.80

 安装

-1.77

 简介

-1.76

 هنری

-1.75

 شهرد

-1.73

 melk

-1.72

 bluse

-1.72

 项目

-1.71

POSITIVE LOGITS

ֶּ

2.06

ִּ

1.94

 السياس

1.92

 солнца

1.92

ּוֹ

1.85

וּ

1.78

ְּ

1.75

ַּ

1.70

此事

1.69

 especí

1.68

Activations Density 0.172%