INDEX

Explanations

The neuron activates on content-bearing tokens (meaningful nouns/verbs/identifiers and keywords) rather than on common function words or punctuation.

New Auto-Interp

Configuration

Prompts (Dashboard)

273,612 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SDI

0.92

︠

0.88

苺

0.87

Ireland

0.86

 आंकड़े

0.85

 numArray

0.85

 पलीनोमियल

0.84

৷

0.84

Prescott

0.84

 Ariz

0.83

POSITIVE LOGITS

0.77

 headspace

0.68

0.67

ૃત

0.64

เ

0.59

 hinter

0.59

 deep

0.59

0.57

0.56

若

0.55

Activations Density 0.000%