INDEX

Explanations

common phrases and their continuations

The neuron strongly activates on central content nouns (key objects or concepts) in a technical or descriptive context.

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

 acknowledging

-1.16

もちろん

-1.15

 giving

-1.10

 incorporating

-1.10

 groundbreaking

-1.08

 leaving

-1.05

to

-1.04

!」

-1.03

 first

-1.02

 perpetually

-1.02

POSITIVE LOGITS

 vijf

1.21

 Reſ

1.20

 plusieurs

1.20

nachweise

1.19

zaSy

1.18

􀀀

1.18

 zeven

1.12

 fabulous

1.11

itting

1.11

 bijna

1.10

Activations Density 0.157%