INDEX

Explanations

This neuron responds to numeric tokens (e.g. years, counts, IDs, and other numbers).

punctuation and "and"

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

lamations

-0.98

 vokal

-0.98

 konkurs

-0.96

kyt

-0.91

 modeller

-0.91

odore

-0.91

 Cataly

-0.91

 الکتر

-0.91

 teks

-0.90

 komentar

-0.90

POSITIVE LOGITS

 regarding

1.41

 Regarding

1.34

Regarding

1.28

 also

1.14

 щодо

1.12

 again

1.12

 Tampoco

1.11

you

1.10

關於

1.05

 dotyczące

1.04

Activations Density 0.010%