INDEX

Explanations

how something is measured

The main thing this neuron does is pick out quantitative measurement and data‐reporting terms (e.g. “measured,” “assessed,” numeric values, scale/index names) in academic text.

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

就开始

-1.34

Transk

-1.27

леку

-1.25

Бил

-1.20

 можете

-1.20

 النبات

-1.18

although

-1.16

 translateY

-1.16

 confronto

-1.15

ׇ

-1.15

POSITIVE LOGITS

 included

1.19

וּ

1.18

 indicator

1.17

ַּ

1.17

וּ

1.16

 кроссовки

1.15

 each

1.13

obatan

1.13

Indicator

1.13

Soorten

1.13

Activations Density 0.053%