INDEX

Explanations

company names and services

The neuron is triggered by corporate names and business‐services jargon (especially IT/consulting/outsourcing firm names and related service terms).

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

 treu

-1.34

 ragazze

-1.34

 geda

-1.30

nowned

-1.29

 помогают

-1.27

琑

-1.25

 personalità

-1.24

 Called

-1.24

狲

-1.24

옙

-1.23

POSITIVE LOGITS

 blusa

1.41

1.32

ญี่ป

1.30

你会

1.29

vább

1.27

主催

1.25

你需要

1.24

 acides

1.23

лся

1.22

 מאוד

1.20

Activations Density 0.023%