INDEX

Explanations

document start phrases

The neuron detects LaTeX package‐import lines (i.e. “\usepackage{…}” commands).

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

 took

-0.91

 Пав

-0.75

 until

-0.74

느

-0.72

 settembre

-0.72

 Agustus

-0.70

爽

-0.69

 does

-0.68

 conseil

-0.68

 atti

-0.67

POSITIVE LOGITS

 comienzan

0.88

erno

0.85

 hábito

0.84

 commencé

0.83

 comenzó

0.80

ワクワク

0.79

 empez

0.78

euvre

0.78

 começou

0.77

 Crum

0.77

Activations Density 0.000%