INDEX

Explanations

to be

This neuron detects phrases expressing assessment or appearance—particularly the “to be” infinitive following verbs like look, sound, or seem (e.g. “looks to be,” “sounds like”).

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

by

-1.73

]);

-1.59

%";

-1.59

！！”

-1.58

ୈ

-1.57

𖥦

-1.55

lorin

-1.55

 What

-1.52

]&

-1.51

”，

-1.50

POSITIVE LOGITS

1.69

 youngsters

1.48

近い

1.42

1.41

不错

1.41

Filosof

1.39

1.38

raszamy

1.35

 leftovers

1.33

 taux

1.30

Activations Density 0.008%