INDEX

Explanations

grilling and barbecuing

This neuron fires on structured specification or feature‐listing text—especially tokens conveying numeric measurements or technical parameters.

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

 pastry

-0.87

 coffees

-0.85

 violin

-0.83

 coffee

-0.81

 oats

-0.81

oat

-0.81

 pastries

-0.81

BATH

-0.80

 Ballroom

-0.80

HIV

-0.79

POSITIVE LOGITS

 grill

3.20

 barbecue

3.14

BBQ

3.05

 grilling

2.91

 grills

2.55

 Grill

2.53

bbq

2.44

 barbe

2.41

grill

2.38

BBQ

2.38

Activations Density 0.047%