INDEX

Explanations

practical explanations and tutorials

The neuron strongly activates on numerical and numeric‐style tokens (e.g. years, multi-digit figures, dates, statistics).

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

()}")

0.55

려는

0.54

 dugg

0.49

噸

0.49

xmax

0.48

 повре

0.47

얗

0.46

 enfermed

0.46

กลาง

0.46

కుని

0.46

POSITIVE LOGITS

 Writing

0.97

 Thesis

0.96

Writing

0.83

Thesis

0.83

 Rehabilitation

0.82

 Neuroscience

0.80

 ChatGPT

0.78

 thesis

0.76

 Drafting

0.76

 Engineering

0.75

Activations Density 0.084%