INDEX

Explanations

James Madison

The neuron is detecting proper personal names (named entities referring to people).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 detonation

0.68

 soliton

0.58

 idempot

0.57

紧

0.56

 harmonics

0.56

INCLUDING

0.56

緊

0.56

 Veget

0.55

Detach

0.55

 RICH

0.55

POSITIVE LOGITS

aient

0.70

ardless

0.64

ごろ

0.63

 zaradi

0.62

 duidelijk

0.61

yll

0.60

Puy

0.59

ian

0.59

avy

0.59

 berbicara

0.57

Activations Density 0.128%