INDEX

Explanations

terms, configuration, presentation, distribution, goals

This neuron detects documentation section headers and instructional phrases in comments (e.g. “Specifies,” “When configured,” “For more details on how to…”).

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

 their

-2.11

it

-1.90

its

-1.81

 first

-1.77

 only

-1.76

as

-1.71

on

-1.62

 this

-1.61

 another

-1.60

an

-1.57

POSITIVE LOGITS

甗

1.43

 diyor

1.38

很是

1.38

ांक

1.34



1.32

 utiliser

1.31

说实话

1.30

臃

1.27

 éviter

1.27

 ardu

1.26

Activations Density 0.068%