INDEX

Explanations

of glory and bidding

The neuron strongly activates on grandiose or authoritative abstract nouns (like “will,” “glory,” “kingdom,” “bidding,” “empire,” etc.).

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

TintMode

-1.84

athed

-1.50

licopter

-1.49

秫

-1.48

还真

-1.47

зыв

-1.47

before

-1.46

佧

-1.46

liśmy

-1.45

協會

-1.45

POSITIVE LOGITS

is

1.63

1.61

er

1.57

 носить

1.40

 oscura

1.34

Our

1.31

 first

1.31

asına

1.30

be

1.29

 subtleties

1.28

Activations Density 0.006%