INDEX

Explanations

conditional logic `if` statements

The neuron is primarily detecting numeric literals (stand-alone numbers) in the code.

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

⪢

-1.03

saa

-0.98

 Mittwoch

-0.98

 sommeil

-0.97

 jonka

-0.97

 épisode

-0.92

taan

-0.90

 rather

-0.90

ㅍ

-0.89

 rocher

-0.88

POSITIVE LOGITS

if

2.23

 excessive

1.28

 even

1.16

 excessively

1.16

 check

1.12

 when

1.05

 intense

1.04

 furthermore

1.02

 vizsg

1.01

 zudem

1.00

Activations Density 0.002%