INDEX

Explanations

file paths

This neuron detects file‐system path components (e.g. absolute directory or repository URLs).

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

dekat

-1.41

ادو

-1.27

 with

-1.25

 motocicleta

-1.24

 яйца

-1.23

szy

-1.22

 digunakan

-1.19

๔

-1.19

 botas

-1.18

 konkurs

-1.18

POSITIVE LOGITS

na

1.41

1.34

ck

1.29

1.26

 worin

1.25

辛い

1.23

ry

1.22

le

1.22

1.21

1.19

Activations Density 0.002%