INDEX

Explanations

can't get something to work

The neuron strongly fires on common simple verbs—especially “get” (including negated forms like “can’t get”) and similar action words such as “stop.”

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

⢪

-1.45

 spesa

-1.44

感じで

-1.44

髕

-1.41

一隻

-1.41

;-

-1.40

libft

-1.39

⸫

-1.38

 sichere

-1.36

 dringend

-1.35

POSITIVE LOGITS

for

1.38

nem

1.25

of

1.23

où

1.19

нка

1.18

 bepaald

1.15

 デイ

1.14

TAKT

1.13

 unplanned

1.13

айт

1.13

Activations Density 0.011%