INDEX

Explanations

First last name pairs

This neuron identifies personal names, especially author bylines and proper‐noun name pairs in 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

 will

-1.67

by

-1.34

czegóły

-1.29

 from

-1.27

as

-1.26

out

-1.23

 sprawy

-1.21

活動を

-1.19

 soldat

-1.16

asible

-1.15

POSITIVE LOGITS

 mismas

1.48

 that

1.45

Also

1.34

But

1.30

 trám

1.25

↵↵

1.24

Kör

1.24

這

1.23

 тогда

1.21

 succede

1.21

Activations Density 0.121%