INDEX

Explanations

along walls, roads, banks

The neuron activates on words naming physical edges or borders (e.g. “bank,” “wall,” “roadside”).

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

 since

-1.48

Since

-1.42

年から

-1.34

If

-1.27

With

-1.20

 from

-1.20

if

-1.20

From

-1.19

𝓇

-1.17

for

-1.15

POSITIVE LOGITS

 entlang

1.50

 вдоль

1.20

UNCIL

1.19

回来了

1.16

oughby

1.14

 ALONG

1.13

 ocur

1.13

ኽ

1.12

succeed

1.12

 capilar

1.12

Activations Density 0.063%