INDEX

Explanations

introductions and scientific notation

This neuron detects mathematical norm notation—specifically the double‐bar “∥ … ∥” constructions indicating vector/function norms.

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

ithe

-1.43

 взаимодей

-1.22

部分を

-1.16

⸫

-1.16

("");

-1.13

 😁

-1.11

部分が

-1.11

OOOO

-1.09

?!?

-1.09

?????

-1.09

POSITIVE LOGITS

 like

1.58

on

1.40

 through

1.38

in

1.37

 very

1.20

＝＝

1.19

 good

1.18

as

1.18

not

1.17

 strangely

1.16

Activations Density 0.004%