INDEX

Explanations

if not

The neuron strongly activates on the negation phrase “if not,” especially on the word “not” in that construction.

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

 there

-1.88

and

-1.81

</td>

-1.81

-1.59

-1.55

te

-1.52

-1.50

-1.45

-1.43

 with

-1.42

POSITIVE LOGITS

 sekarang

1.71

ння

1.62

 maintenant

1.59

cadeau

1.58

jours

1.55

 cette

1.55

 Üniversitesi

1.52

 immort

1.52

ẃ

1.50

 serce

1.49

Activations Density 0.007%