INDEX

Explanations

actor, actress, firearm, river, mountain

The neuron is highly activated by unusually long tokens—that is, multi-syllable or “long” words.

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

 zeven

-1.41

 estadounid

-1.37

 meinem

-1.34

儛

-1.33

さみ

-1.30

 listView

-1.29

 soixante

-1.27

 soyez

-1.27

 hemd

-1.25

 mogli

-1.24

POSITIVE LOGITS

 всегда

1.45

 который

1.38

 with

1.35

它的

1.24

its

1.23

lineno

1.23

万年

1.19

发布时间

1.19

袂

1.18

cellona

1.17

Activations Density 0.175%