INDEX

Explanations

code syntax and notation

The neuron strongly activates on command-style verbs (imperatives) appearing as standalone UI actions or headings (e.g. “Ask,” “Save”).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

hint

0.60

vect

0.59

client

0.57

두

0.57

Decimal

0.56

otia

0.55

col

0.55

hoa

0.54

중

0.54

POSITIVE LOGITS

 pallets

0.68

 acetonitrile

0.67

 mele

0.67

 రూ

0.66

 tablas

0.65

 kneel

0.64

 tapas

0.64

 raat

0.63

 dets

0.63

 هاي

0.63

Activations Density 0.000%