INDEX

Explanations

for all / ∀

This neuron fires on formal mathematical formula notation—especially set‐builder or similar expressions featuring symbols like “∈”, “|” (or “\mid”), “=”, and quantifiers (e.g. “∀”).

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

 mientras

-0.90

 Anreg

-0.88

 våre

-0.88

жмите

-0.87

hassee

-0.86

 allerede

-0.85

varer

-0.84

 polskiej

-0.84

たちも

-0.84

 También

-0.82

POSITIVE LOGITS

forall

4.13

for

3.63

∀

2.84

ForAll

2.27

 לכל

2.11

 every

1.77

For

1.71

 لكل

1.70

for

1.68

all

1.65

Activations Density 0.045%