INDEX

Explanations

The neuron primarily fires on specialized scientific measurement terms and quantitative parameter labels (e.g., critical current density Jc, critical temperature Tc, “superconducting,” etc.).

superconductivity, magnetism, and insulators

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

if

-1.42

幸好

-1.41

 sedu

-1.36

𝗼

-1.35

ETS

-1.31

şiv

-1.30

 sirve

-1.30

 effettu

-1.28

 sullen

-1.27

SuccessListener

-1.27

POSITIVE LOGITS

not

1.64

on

1.47

but

1.40

 both

1.36

 this

1.31

Público

1.28

all

1.28

 only

1.27

pergillus

1.27

 like

1.26

Activations Density 0.170%