INDEX

Explanations

the same

The neuron activates on numeric expressions and quantitative measurements (numbers, units, counts, or measurement-related tokens).

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

 just

-1.70

⅘

-1.43

砹

-1.34

 ilyen

-1.33

 vraiment

-1.32

 seiner

-1.30

 bahkan

-1.29

 seulement

-1.27

女生

-1.27

 aussi

-1.26

POSITIVE LOGITS

 that

2.42

 than

1.93

 they

1.57

>${

1.31

りましたが

1.27

></

1.19

んですけど

1.19

(=)

1.16

[{\

1.15

':{'

1.14

Activations Density 0.105%