INDEX

Explanations

technical specifications and identifiers

The neuron strongly activates on multi‐digit numeric sequences (e.g. version numbers, file path IDs, or other long runs of digits).

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

Carex

-1.15

so

-1.08

or

-1.07

 while

-1.05

　　——

-1.01

Porte

-1.00

ؤول

-0.98

呻

-0.97

 Often

-0.95

жал

-0.95

POSITIVE LOGITS

 cuando

1.33

その

1.13

 Betrach

1.11

استفاده

1.07

さらに

1.07

 Stoffe

1.05

ですね

1.02

 languid

0.99

 gesuchten

0.98

 quando

0.97

Activations Density 0.116%