INDEX

Explanations

justin followed by person's name

The neuron responds to capitalized tokens—especially proper names, titles, or other uppercase‐initial words.

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

fos

-0.75

QObject

-0.72

包裹

-0.72

乾

-0.71

parseDouble

-0.71

 negros

-0.70

 houses

-0.70

 trans

-0.69

andle

-0.69

НИЙ

-0.69

POSITIVE LOGITS

Bie

1.32

 Justin

1.31

 Bieber

1.29

Bie

1.25

Justin

1.21

 Stratford

1.21

 Hailey

1.14

Purpose

1.12

 Purpose

1.11

 belieb

1.07

Activations Density 0.009%