INDEX

Explanations

structural elements: percentages, parentheses, pipe separators

The neuron fires on numeric tokens—counts, measurements, and statistical values—especially those found in tables or indices.

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

 However

-1.58

 These

-1.38

it

-1.38

 Though

-1.25

providedIn

-1.20

 There

-1.20

the

-1.20

 Although

-1.18

The

-1.16

It

-1.16

POSITIVE LOGITS

</h2>

1.29

erobic

1.29

 katk

1.16

Giới

1.16

 viime

1.13

letics

1.12

极其

1.12

 angelegt

1.10

极为

1.10

 maksi

1.10

Activations Density 0.010%