INDEX

Explanations

The neuron strongly activates on sequences of capitalized personal names (e.g. author lists in academic citations).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Органи

0.30

 Пла

0.29

 Tecnología

0.28

 Лі

0.28

 예술

0.28

 Фа

0.27

 Ком

0.27

ío

0.27

 Хар

0.27

 Agustín

0.27

POSITIVE LOGITS

0.27

cok

0.27

<0x0D>

0.25

tag

0.25

 tonic

0.24

 tags

0.23

 arit

0.23

 variable

0.22

 monotonic

0.22

Activations Density 1.986%