INDEX

Explanations

variables like i, b, m, total, 2, cap

The neuron fires on user‐provided code identifiers—i.e. variable names in code snippets.

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

 kabát

-1.37

kó

-1.34

 euró

-1.22

 kré

-1.21

the

-1.20

 '../../

-1.15

nowu

-1.13

 defekt

-1.10

 technik

-1.09

 spannend

-1.04

POSITIVE LOGITS

 함께

1.24

 이야

1.16

링

1.15

 rump

1.09

케

1.04

뉴

1.04

 גם

1.02

書く

1.02

宕

1.02

 아트

1.01

Activations Density 0.008%