INDEX

Explanations

This neuron specifically detects occurrences of the syllable “la” (as in repeated “la la la” vocalizations).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 socalled

1.08

 wellknown

0.91

ल्लिंग

0.84

 followup

0.80

merksamkeit

0.79

___

0.78

িয়৷

0.71

 ______

0.71

 nondegenerate

0.69

 longterm

0.69

POSITIVE LOGITS

3.97

-...

2.50

-$

2.31

-(

2.28

-'

2.22

־

2.22

-[

2.17

-{

2.16

‑

2.16

2.14

Activations Density 1.872%