INDEX

Explanations

rounded corners

The neuron fires on mentions of “corner” (e.g. inner corner, outer corner, rounded corners) in instructional or descriptive contexts.

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

it

-1.34

 papild

-1.34

 veiks

-1.20

 people

-1.14

 viena

-1.11

蕯

-1.06

 pazi

-1.06

攬

-1.04

 pribadi

-1.02

雍

-1.00

POSITIVE LOGITS

esomeness

1.39

 które

1.30

 derni

1.29

руются

1.29

 "'");

1.23

 které

1.23

 ")");

1.20

 المناطق

1.20

 ")";

1.19

are

1.17

Activations Density 0.038%