INDEX

Explanations

code snippets and programming constructs

This neuron fires on code identifiers ending in typical noun-forming suffixes like “-tion” or “Type” (e.g. section, Type, getNodeType).

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

 informée

-1.23

⿻

-1.10

 FEBR

-1.08

ضاف

-1.07

┈

-1.04

 ingat

-1.04

 sukses

-1.04

 mystérie

-1.03

Semua

-1.02

ଃ

-1.02

POSITIVE LOGITS

 stwor

1.10

 different

1.08

 unterschied

1.07

 normal

1.03

することが

1.00

0.97

sh

0.96

we

0.96

できて

0.96

 convict

0.96

Activations Density 0.000%