INDEX

Explanations

code block structure

The neuron fires on module‐loading or dependency declarations (e.g. import, require, use, plan/skip checks), i.e. lines that bring in external libraries or set up the test environment.

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

if

-1.18

 those

-1.12

これまで

-0.95

 while

-0.93

ඛ

-0.93

 that

-0.93

واصل

-0.92

編曲

-0.92

 kän

-0.91

had

-0.90

POSITIVE LOGITS

 test

1.38

 también

1.18

测试

1.10

ÈRE

1.07

ọ

1.05

 testing

1.05

 mock

1.05

test

1.04

johtaja

1.02

illères

1.00

Activations Density 0.040%