INDEX

Explanations

compare, explain, describe, summarize

This neuron specifically detects the imperative phrase “Distinguish between” (i.e. instructions asking to distinguish/compare two things).

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

during

-1.52

 них

-1.45

 durante

-1.29

 poivre

-1.20

 them

-1.20

 ними

-1.19

cessed

-1.16

-1.14

 aplicativo

-1.13

POSITIVE LOGITS

and

1.95

how

1.52

Dá

1.26

 required

1.24

 bereits

1.20

 different

1.20

 necessary

1.20

皇子

1.18

stuk

1.18

fromString

1.17

Activations Density 0.119%