INDEX

Explanations

explaining unique things and concepts

The neuron fires on quantitative, data‐heavy tokens—especially numbers, measurements, percentages, and related methodological terms—in scientific or technical contexts.

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

Pä

-1.57

Maintenant

-1.56

 perpé

-1.55

 cotone

-1.55

 migli

-1.50

 réserver

-1.49

优秀的

-1.47

 estadounid

-1.44

有一個

-1.43

Cliquez

-1.43

POSITIVE LOGITS

for

1.91

1.48

 sieben

1.26

!!!!!!

1.21

此刻

1.14

 zoals

1.14

DOI

1.14

さらに

1.12

!!!!!!!!

1.11

!!!!!!!

1.11

Activations Density 0.021%