INDEX

Explanations

HTML class attributes

The neuron is primarily responding to numeric literals (especially floating‐point numbers) in math expressions and code.

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

to

-1.63

 такі

-1.41

ebly

-1.39

少しでも

-1.38

謌

-1.36

合わせた

-1.35

以外にも

-1.30

вање

-1.28

istungen

-1.27

nienie

-1.27

POSITIVE LOGITS

2.02

and

1.84

allez

1.71

1.48

 ویکی

1.44

，「

1.38

kleid

1.36

</u>

1.33

1.28

1.25

Activations Density 0.268%