INDEX

Explanations

mathematical and computational terms

This neuron detects discourse cues and instructional/explanatory markers—words that introduce examples, notes, steps, or descriptions in explanatory text.

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

 Hospice

0.88

testimonials

0.83

 Interfaith

0.83

 Bazaar

0.82

 eCommerce

0.80

 Gaming

0.80

LGBT

0.80

Gaming

0.79

 지속

0.79

 اکثر

0.78

POSITIVE LOGITS

 algebraic

1.48

 algebraically

1.37

 algebra

1.27

 computations

1.23

 computation

1.23

 subscripts

1.20

 intuition

1.19

 automorphism

1.19

 mathemat

1.16

 mathematical

1.16

Activations Density 0.182%