INDEX

Explanations

The neuron fires strongly on content-bearing tokens—nouns, main verbs, adjectives, and other informative words—while ignoring common function words.

manufacture, production, built, 50s, 1958

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

now

-1.57

🪛

-1.45

 already

-1.41

 currently

-1.40

ıyoruz

-1.37

regelen

-1.30

 recently

-1.24

new

-1.23

بيرة

-1.22

 explorar

-1.21

POSITIVE LOGITS

of

1.34

later

1.22

 Août

1.20

 suivants

1.19

 поиск

1.17

 later

1.15

𝐘

1.15

 سریال

1.13

 Février

1.13

 Décembre

1.13

Activations Density 0.045%