INDEX

Explanations

RCS file versions

The neuron strongly responds to short all-caps alphabetic tokens (acronyms or uppercase codes) in code comments.

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

and

-2.97

dáms

-1.77

抬手

-1.69

蒞

-1.64

 bicicletas

-1.62

ついた

-1.59

入った

-1.58

agosto

-1.56

 áng

-1.54

unuz

-1.53

POSITIVE LOGITS

 tidak

1.77

 sconto

1.75

 età

1.72

謌

1.70

lemma

1.68

зации

1.66

ries

1.66

 this

1.65

盌

1.63

MÁS

1.62

Activations Density 0.001%