INDEX

Explanations

with opening files or urls

The neuron is specialized to detect attribute‐access patterns in code (i.e. the “dot” operator connecting an object like “self” to its attribute or method).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 when

-1.36

by

-1.30

 their

-1.27

を持っている

-1.13

 will

-1.09

 would

-1.05

had

-1.03

 calça

-1.02

 waarop

-1.01

 quando

-0.98

POSITIVE LOGITS

():

1.45

 namens

1.17

Ḥ

1.16

mitian

1.07

 качестве

1.04

 scipy

1.02

एस

1.02

):

1.02

roffenen

1.01

]:

0.98

Activations Density 0.004%