INDEX

Explanations

The neuron strongly responds to code tokens that introduce named arguments or attributes (e.g. “name=…”, “related_name=…”, “Key=…”) in programming snippets.

questions and definitions

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

 already

-1.20

 with

-1.11

 this

-1.03

已是

-1.03

 bereits

-1.01

 what

-0.95

ších

-0.94

 those

-0.92

who

-0.91

 уже

-0.91

POSITIVE LOGITS

cuando

1.17

蜕

1.16

Ingredi

1.13

hagic

1.08

ukų

1.05

HOPE

1.05

他們的

1.04

↑↑↑

1.04

Hasil

1.03

 revamped

1.03

Activations Density 0.000%