INDEX

Explanations

вопросительные предложения

The neuron activates on numeric tokens or explicit quantitative information (e.g. measurements, counts, dates, file‐sizes).

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

Nombre

1.17

cual

1.16

ÍA

1.15

respuesta

1.12

asign

1.10

 niveles

1.09

ensión

1.05

Alumno

1.05

ían

1.05

oxygène

1.05

POSITIVE LOGITS

 при

1.84

с

1.75

 про

1.72

і

1.65

 по

1.64

в

1.61

 до

1.55

 со

1.54

 от

1.53

 не

1.52

Activations Density 0.493%