INDEX

Explanations

laminin, laminitis, Lagos, latent

The neuron detects occurrences of the substring “lamin” (e.g. in “laminin,” “lamina,” etc.).

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

 ถ้า

-1.34

 不锈钢

-1.34

 attivo

-1.34

੮

-1.32

 industriale

-1.31

铪

-1.30

瓘

-1.29

XICO

-1.27

 parlar

-1.27

 sentimento

-1.27

POSITIVE LOGITS

 как

1.26

</b>

1.25

(){

1.17

ev

1.15

 credence

1.10

 pretend

1.09

PA

1.09

izations

1.07

 kicks

1.06

1.05

Activations Density 0.057%