INDEX

Explanations

code block endings or continuations

The neuron activates on code‐block boundary markers—tokens that close or end blocks (e.g. “}”, “catch”, “end”, “?>”, “</script>”).

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

-1.57

:"",

-1.48

및

-1.42

quelles

-1.41

Доку

-1.36

๕

-1.36

ții

-1.30

arkhand

-1.27

༺

-1.26

Quaternion

-1.25

POSITIVE LOGITS

If

1.42

if

1.39

 that

1.37

for

1.37

 when

1.36

It

1.34

use

1.33

 because

1.30

1.24

1.23

Activations Density 0.020%