INDEX

Explanations

technical code initialization and creation

The neuron is keyed to function‐ or method‐call sites in code—i.e. it fires on identifiers immediately followed by “(” (and on the matching “{“, “}” or other punctuation around those calls).

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

 garagem

-0.99

räder

-0.92

creating

-0.91

 官网

-0.90

 grazia

-0.89

rätt

-0.89

 rengi

-0.89

 机器

-0.88

צל

-0.87

绿

-0.86

POSITIVE LOGITS

 from

1.33

 décédé

1.01

..

0.95

new

0.94

ング

0.89

焚

0.88

 ....

0.84

 trägt

0.84

 .....

0.82

 besonders

0.82

Activations Density 0.019%