INDEX

Explanations

verbs leading to states or actions

The neuron strongly activates on descriptive action verbs—especially past‐tense or participial forms (e.g. “drawled,” “wracked,” “spreading,” “cradled,” “grasped”) that convey physical motion or sensory experience.

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

polazione

-1.20

ními

-1.14

};

-1.07

you

-1.05

nione

-1.04

 württemberg

-1.03

 preved

-1.03

Then

-1.03

ativní

-1.02

indelijk

-1.02

POSITIVE LOGITS

of

1.55

he

1.16

ական

1.15

 dosta

1.08

previously

1.07

anteed

1.05

 ansehen

1.05

 DONE

1.02

versial

1.00

 spécialiste

1.00

Activations Density 0.033%