INDEX

Explanations

via simple, mechanical fasteners, #! usr

The neuron activates on technical equation tokens—Greek-letter variables (μ, γ), field symbols (E, H), and related mathematical symbols—i.e. components of scientific formulae.

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

 تساعد

-1.51

 like

-1.35

 that

-1.30

or

-1.27

 help

-1.27

its

-1.23

()

-1.17

ٗ

-1.16

 правительства

-1.16

 mentre

-1.16

POSITIVE LOGITS

鸨

1.27

our

1.25

 opět

1.21

styrelsen

1.20

хіво

1.15

 châ

1.15

 naše

1.13

 μας

1.11

 individuelles

1.09

 saveurs

1.09

Activations Density 0.008%