INDEX

Explanations

lower height, flutes, axis, density, toe

The neuron activates on technical specification language detailing component placement or configuration (e.g. “legs can be ‘outset’ at the corners,” “grille for maximum heating,” “lower to give more airflow,” etc.).

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

镂

-1.15

 thier

-1.12

automne

-1.10

诙

-1.08

 tuyau

-1.06

Рейтинг

-1.06

說

-1.05

 including

-1.04

 soutien

-1.02

无论是

-1.02

POSITIVE LOGITS

so

1.67

to

1.51

 afin

1.45

 чтобы

1.44

 sehingga

1.38

เพื่อ

1.30

 כדי

1.25

hauled

1.25

 permite

1.22

 để

1.16

Activations Density 0.060%