INDEX

Explanations

The neuron activates on technical descriptions of data‐integrity processes—especially mentions of checksums, detecting and correcting errors, and related storage‐system jargon.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

0.86

0.80

0.73

0.72

Tee

0.72

0.71

 forman

0.70

(.

0.70

kao

0.69

 goede

0.69

POSITIVE LOGITS

 whenever

0.89

whenever

0.89

每一次

0.88

すべての

0.88

 невероят

0.88

 যখন

0.87

جميع

0.86

 når

0.86

Após

0.85

每個

0.85

Activations Density 0.122%