INDEX

Explanations

introducing someone or something

The neuron fires on brief speaker‐introduction phrases—e.g. “This is X speaking” or “X here”—marking when someone names themself at the start of dialogue.

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

我们的

-1.30

 bahkan

-1.26

 recentes

-1.24

 EXERC

-1.24

 wiele

-1.24

 навіть

-1.23

 provinsi

-1.22

 betreffende

-1.21

 certamente

-1.19

 جلسه

-1.18

POSITIVE LOGITS

ReferenceEquals

1.36

 representa

1.34

瘩

1.25

out

1.24

for

1.23

ྃ

1.21

ipolar

1.20

 که

1.20

 from

1.19

 after

1.18

Activations Density 0.026%