INDEX

Explanations

The neuron detects occurrences of the token “com” in Java/Kotlin package or import statements.

java package com

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

in

-2.14

on

-1.91

all

-1.74

 both

-1.67

 within

-1.65

any

-1.63

 their

-1.48

 being

-1.45

 have

-1.43

out

-1.41

POSITIVE LOGITS

 tutkim

1.33

 gedachten

1.30

為に

1.30

ほら

1.27

↑↑↑

1.27

 dette

1.26

是一

1.22

 området

1.21

 horrid

1.21

Activations Density 0.010%