INDEX

Explanations

mathematical equations and variables

This neuron fires on LaTeX‐style math expressions (integral signs, “\” commands and surrounding equation tokens).

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

eturn

-1.34

쨩

-1.30

𓆏

-1.25

 konkret

-1.23

為に

-1.23

ܙ

-1.17

 économies

-1.16

 séjours

-1.15

:"

-1.13

跸

-1.13

POSITIVE LOGITS

of

1.55

 other

1.52

now

1.44

 after

1.43

 others

1.41

by

1.34

 because

1.25

 very

1.23

at

1.20

 showing

1.18

Activations Density 0.065%