INDEX

Explanations

concepts based on research methods

The neuron activates on content‐bearing technical nouns that name or label research methods, tools, or study components (e.g. “calculations,” “effects,” “study,” “approach,” “tool,” “algorithm,” “parameters,” 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

its

-1.45

it

-1.42

用

-1.24

 législation

-1.18

성

-1.17

 его

-1.14

把它

-1.13

 רבה

-1.13

 zufolge

-1.13

 Pandey

-1.13

POSITIVE LOGITS

jeux

1.32

่วน

1.27

avril

1.17

inglés

1.12

 namely

1.09

 based

1.08

 dimana

1.08

 フロア

1.06

Livraison

1.06

やかに

1.06

Activations Density 0.294%