INDEX

Explanations

not be

The neuron detects negation phrases, especially the word “not” (often in constructions like “may not,” “might not,” “not be,” 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

 Gaulle

-1.62

'},

-1.59

ține

-1.56

 ราคา

-1.52

 ſou

-1.51

)}{\

-1.51

 qualquer

-1.50

 produisent

-1.48

 vivent

-1.47

〬

-1.42

POSITIVE LOGITS

 have

1.94

 even

1.65

be

1.59

未必

1.48

 well

1.45

1.44

But

1.44

 also

1.41

However

1.36

1.35

Activations Density 0.020%