INDEX

Explanations

detecting questions

This neuron detects interrogative constructions—phrases forming a question (e.g., “Is there…?”, “How was…?”, “Will it…?”).

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

市场的

-0.99

скра

-0.98

 vuelos

-0.94

ые

-0.92

eliminar

-0.90

 küche

-0.86

 zarówno

-0.86

ETRY

-0.86

从事

-0.85

LOAT

-0.84

POSITIVE LOGITS

1.16

you

1.06

 ikiwa

1.05

1.04

we

1.03

0.95

тельностью

0.94

 responsabilidad

0.94

cembre

0.93

 вдруг

0.92

Activations Density 0.065%