INDEX

Explanations

either or

The neuron activates primarily on numeric literal tokens (i.e. integers or floats) in the 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

-1.95

);

-1.41

}$

-1.39

 most

-1.36

 $^{\

-1.34

-1.33

$^{

-1.33

 экономики

-1.31

 senhora

-1.26

...\

-1.23

POSITIVE LOGITS

 either

1.69

either

1.64

,’

1.56

 beina

1.51

⑳

1.47

嚒

1.46

ܜ

1.43

 alcun

1.42

瞜

1.42

GEBUR

1.41

Activations Density 0.015%