INDEX

Explanations

later events

The neuron detects temporal shift markers—words like “Later,” date or time expressions, and similar cues that signal a change or reference in chronology.

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

${

1.18

<?

1.07

1.04

 Grâce

1.02

ं

1.00

我现在

0.98

 mailing

0.96

Đây

0.96

τούν

0.96

POSITIVE LOGITS

್

1.08

َ

0.92

ر

0.90

د

0.88

 metais

0.87

 dese

0.85

pore

0.85

ی

0.85

 মুখোমুখি

0.83

 stages

0.82

Activations Density 0.048%