INDEX

Explanations

The neuron strongly activates on “available in” and similar constructions that introduce lists of options, variants, or offerings (e.g. finishes, colours, styles).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 물질

0.60

 деятельность

0.59

 ശരീ

0.58

 परिस्थिती

0.57

 aquilo

0.57

Definition

0.56

approach

0.56

 estrut

0.55

 tevé

0.55

 estrutura

0.55

POSITIVE LOGITS

 sizes

1.73

 Sizes

1.47

sizes

1.45

 tamaños

1.39

Sizes

1.38

 variants

1.27

 versions

1.26

 tailles

1.26

 colors

1.21

 flavors

1.21

Activations Density 0.808%