INDEX

Explanations

ended, culminating, terminating

This neuron detects verbs or verb phrases that indicate something concluding or culminating in a particular outcome (e.g. “ended in,” “culminate in,” “terminating in”).

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

。

-1.73

 from

-1.73

of

-1.68

новение

-1.60

'';

-1.45

’’

-1.41

not

-1.41

sistence

-1.39

仍

-1.38

 still

-1.36

POSITIVE LOGITS

 consommate

1.54

ハッピー

1.48

↓

1.45

 jiné

1.45

 paie

1.44

 souver

1.44

怼

1.41

ﮦ

1.41

お待ち

1.40

 风格

1.39

Activations Density 0.052%