INDEX

Explanations

code constructs and identifiers

The neuron fires on the “:” prefix of symbol literals (e.g. the `(:name)` or `:symbol` syntax) in Ruby code.

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

It

-1.80

 มัน

-1.64

 Перейти

-1.63

 Incidentally

-1.57



-1.56

 hauptsächlich

-1.53

akyti

-1.52

鉿

-1.52

corollary

-1.49

Jura

-1.47

POSITIVE LOGITS

end

1.89

1.82

after

1.67

1.63

1.62

 czyli

1.60

 three

1.58

self

1.56

all

1.54

and

1.52

Activations Density 0.008%