INDEX

Explanations

This neuron detects past-tense verbs that announce something has ended, disappeared, or returned (e.g., “gone,” “over,” “back”).

gone, over, dead

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

︲

-1.38

 pájaro

-1.33

 geändert

-1.32

 erwähnten

-1.30

有时候

-1.30

 việc

-1.29

</em>

-1.27

of

-1.26

饃

-1.25

这样

-1.24

POSITIVE LOGITS

 titul

1.40

appellant

1.37

elections

1.35

Ḥ

1.33

 regresó

1.28

 with

1.27

😦

1.27

Ufficio

1.26

éfonos

1.25

 akad

1.25

Activations Density 0.021%