INDEX

Explanations

for loop iteration

The neuron fires on loop‐index variables (e.g. “i”, “j”, “k”, “x”) when they appear in for‐loop headers or index assignments.

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

-1.51

 laim

-1.27

 amplit

-1.20

of

-1.17

our

-1.17

 great

-1.16

 vieta

-1.14

new

-1.14

leč

-1.11

 alimentare

-1.09

POSITIVE LOGITS

 extraordinaire

1.26

 το

1.23

onu

1.20

 света

1.20

pping

1.20

 остальных

1.19

樯

1.19

cing

1.19

dding

1.18

ulates

1.16

Activations Density 0.010%