INDEX

Explanations

a series of events

The neuron activates strongly on occurrences of the phrase “series of …,” i.e. when a text mentions “series of” followed by some items.

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

始めて

-1.36

 Erster

-1.29

zzlies

-1.23

atiles

-1.22

írito

-1.21

vedať

-1.20

时的

-1.18

 entirety

-1.17

 anyone

-1.17

<bos>

-1.17

POSITIVE LOGITS

 olika

1.43

棐

1.34

ógicas

1.23

髎

1.22

eterangan

1.21

˃

1.19

reuil

1.16

 surla

1.13

 vasos

1.12

夏季

1.10

Activations Density 0.039%