INDEX

Explanations

The neuron primarily picks out numeric tokens with decimal points (i.e. floating-point numbers).

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

olyb

0.79

widgetTo

0.78

进而

0.77

 automobiles

0.76

 собою

0.75

 supposition

0.75

UID

0.73

marshalN

0.73

ोले

0.73

 სას

0.73

POSITIVE LOGITS

 facil

0.77

ور

0.76

 predictor

0.73

ți

0.70

usi

0.68

дым

0.66

acán

0.66

洛

0.66

ang

0.64

 harrowing

0.64

Activations Density 0.000%