INDEX

Explanations

superscript and subscript numbers

This neuron activates on tokens containing a superscript “2” (e.g. squared units like km², mm², cm² or chemical/abbreviation form like A₂, I²).

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

-2.67

 estadounidense

-2.61

-2.55

ቺ

-2.53

彞

-2.45

的

-2.44

</b>

-2.39

 básicamente

-2.36

纔

-2.36

洇

-2.22

POSITIVE LOGITS

횃

3.00

."

2.58

闥

2.47

𐑂

2.33

σουν

2.22

 évolution

2.20

买了

2.17

")

2.14

翃

2.14

ของ

2.13

Activations Density 0.002%