INDEX

Explanations

positions, poses, stances, postures

This neuron responds strongly to words naming specific physical stances or poses (e.g. positions, pose, stance, posture).

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

 only

-1.84

ve

-1.37

 them

-1.34

 experience

-1.29

却不

-1.28

-1.23

 there

-1.23

it

-1.23

 wtedy

-1.22

if

-1.20

POSITIVE LOGITS

</b>

1.72

</h5>

1.55

 motifs

1.39

()]);

1.38

 poses

1.38

coes

1.37

]-->

1.34

 postures

1.34

Външни

1.33

怱

1.32

Activations Density 0.078%