INDEX

Explanations

words beginning with ma

The neuron fires on the letter‐sequence “ma” (in any case) at the start of tokens, i.e. it spots “ma…” prefixes such as in gene/protein names or “MA” abbreviations.

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

 Perceptions

-0.97

 faiths

-0.86

 snows

-0.84

 définiti

-0.81

 fortæller

-0.81

菊

-0.80

 vegetarians

-0.79

 Attitudes

-0.77

liefert

-0.77

illère

-0.75

POSITIVE LOGITS

ma

0.96

Ma

0.95

马

0.94

MA

0.92

Ma

0.92

endorong

0.86

 Ма

0.82

argmax

0.80

 Makes

0.75

 este

0.75

Activations Density 0.055%