INDEX

Explanations

depends on what

The neuron responds to words and markers that signal variability or dependency (e.g. “depends,” “varying,” “depending,” “typically,” plus related discourse connectors).

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

 desgra

-1.07

∡

-1.01

eterangan

-0.96

これは

-0.92

臃

-0.91

inės

-0.90

 tämän

-0.90

 Applies

-0.89

 الأخرى

-0.88

odine

-0.88

POSITIVE LOGITS

 nonetheless

1.30

 generally

1.30

 nevertheless

1.25

 comunque

1.22

 overall

1.20

for

1.16

Nevertheless

1.14

Some

1.11

Generally

1.08

if

1.07

Activations Density 0.037%