INDEX

Explanations

structure and conjunctions

The neuron primarily activates on numeric tokens (i.e., sequences of digits or numbers).

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

 dovr

-1.47

-1.43

عية

-1.25

 délais

-1.23

 garcía

-1.23

 ados

-1.23

tront

-1.23

鹟

-1.20

länd

-1.19

POSITIVE LOGITS

 like

1.48

 suoi

1.28

 или

1.27

or

1.25

 even

1.25

you

1.24

."

1.23

 with

1.21

 его

1.21

1.20

Activations Density 0.280%