INDEX

Explanations

UI group or grid elements

The neuron fires on programming‐language tokens (identifiers, keywords, method names, symbols) rather than ordinary English words.

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

onor

-0.93

orice

-0.92

Oise

-0.82

jina

-0.75

livia

-0.75

^*)

-0.74

srs

-0.73

informatics

-0.73

>`;

-0.72

umy

-0.72

POSITIVE LOGITS

group

0.89

 group

0.84

gc

0.79



0.78

 grid

0.78

picture

0.77

 ribbon

0.76

 simple

0.76

 группе

0.75

GC

0.74

Activations Density 0.004%