INDEX

Explanations

asking about past experience

This neuron detects question-opening phrases like “Have you…?”, “Has anyone…?”, or “Have you tried/considered…?” that invite suggestions or feedback.

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

お疲れ様でした

-1.63

このまま

-1.34

 betyd

-1.27

 некоторых

-1.23

 distante

-1.20

幾個

-1.20

}]

-1.20

phère

-1.16

どうやら

-1.14

濊

-1.14

POSITIVE LOGITS

at

1.86

 other

1.83

to

1.60

 then

1.57

 there

1.53

 also

1.52

new

1.51

 Opiniones

1.34

 allí

1.33

the

1.30

Activations Density 0.023%