INDEX

Explanations

names and roles of people

This neuron detects named-entity tokens, especially personal names, nationalities, and professional titles.

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

 незавер

-1.27

 those

-1.23

 from

-1.22

 some

-1.19

 comprimidos

-1.11

 there

-1.08

 personalità

-1.08

Симпто

-1.07

 treba

-1.05

 mentes

-1.05

POSITIVE LOGITS

who

1.34

</h3>

1.16

 surtout

1.04

already

1.02

=[];

1.00

{{--

0.98

 มา

0.98

 soprattutto

0.96

";

0.96

</h1>

0.96

Activations Density 0.221%