INDEX

Explanations

variable assignments and data structures

The neuron strongly activates on numeric tokens (i.e. isolated numbers or digits) anywhere in the text.

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

来週

-1.33

他们

-1.23

 outras

-1.21

 verwal

-1.18

 versões

-1.16

⸰

-1.16

และ

-1.16

あんな

-1.13

 caixas

-1.12

あー

-1.11

POSITIVE LOGITS

izoen

1.25

henswürdigkeiten

1.08

富有

1.07

 even

1.05

Răsp

1.05

{;

1.02

one

1.02

 кв

1.00

'];

1.00

muy

1.00

Activations Density 0.017%